The Impacts of Quantum Computing On Insurance: February 2021

The impacts of
quantum computing
on insurance
February 2021
The impacts of quantum computing on insurance, 2021
Contents
From theory to reality
The (not so) basics of quantum computing
Applications
Quantum computing’s threat to cyber security
The quantum landscape
Insurance impacts – navigating the quantum realm
Moving forward
From theory to
reality
Quantum computers within reach

Insight
Quantum computing harnesses the quantum mechanical Quantum computing has been around as a theoretical
properties of very small objects, such as superposition and concept since the 1980s, but has only progressed to a In the field of quantum computing, the point at
entanglement, to solve problems that are beyond the tangible reality more recently, with major developments which a programmable quantum device can
reach of classical computing. There is a limit to how much and breakthroughs in hardware and software capabilities.3 perform a task that a classical computer
more powerful our current computers can become. In 1994, Peter Shor introduced the first ever useful cannot perform in a feasible timescale, is
Moore’s Law, which sees computing power double roughly quantum algorithm, which if implemented, could break referred to as quantum supremacy.
every two years, is nearing its limit, due to physical commonly used encryption schemes such as RSA. Shor’s
constraints involved in the further miniaturisation of algorithm, capable of solving mathematical problems In October 2019, Google achieved quantum
transistor chips. Additionally, the speedup in computing underpinning many current cryptograms, which are supremacy for the first time in history. They
offered by parallelization is limited by Amdahl’s law.1 impossibly difficult to solve using classical computing, is claimed that their 54-qubit processor,
Therefore, solving and optimising multi-variable, real-world what ignited widespread interest in actually building the Sycamore, performed a task in 200 seconds,
problems that necessitate the manipulation of large hardware that could support such algorithms.4 However, which would have taken a state-of-the-art
datasets, requires an entirely new paradigm. only in 2019, with Google reaching quantum supremacy, supercomputer 10,000 years.6
was it shown that a quantum computer could actually
Unlike classical computers, which require a two-fold solve a specific problem faster than a classical computer.
increase in transistors to double in power, quantum IBM has since disputed Google’s claim of
computers double in power by the mere addition of one Quantum computing today is more of an engineering quantum supremacy, by suggesting that an
qubit - a quantum bit.2 This means, quantum computers problem than a theoretical one. The current development improved supercomputing technique could
have the potential to deliver significant benefits to many state of quantum computing technology is comparable to theoretically perform the task in just 2.5 days
industries by solving those optimisation, simulation and when classical computers were still using vacuum tubes, (yet a proof of this theoretical technique
machine learning problems, which would otherwise take before their switch to transistors.5 Complex hardware remains to be seen). The clash between
classical computers timescales ranging from 1000s of challenges and a shortage of talent in software these two giants is indicative of the fierce
years to the lifetime of the universe. However, this development, means that a fully-fledged, commercially competition in the private sector to gain
increase in computational power might turn out to be a available quantum computer might still be more than a dominance in this exciting new field.7
double edged sword – encryption systems that safeguard decade away. Nevertheless, the opportunity for
most of our digital communications, and are designed to businesses to take advantage of quantum capabilities Regardless of IBM’s challenge, and despite
be computationally intractable to crack using classical through API based cloud offerings, and discover which of the fact that the task that Sycamore
computing, are potentially vulnerable to the speedups their needs quantum computers could eventually serve, is performed has no real-world application,
offered by quantum algorithms. already within reach. Google’s achievement remains ground-
breaking and has brought the reality of
quantum computing within closer reach.
Source: (1) Chojecki, 2019 (2) IBM Institute for Business Value, 2019 (3) McKinsey Quarterly, 2020 (4) Deloitte University Press, 2017 (5) RAND Corporation, 2020 (6) Arute, 2019 (7) Nature, 2019
The (not so)
basics of quantum
computing
The power of qubits

In classical computing, data is represented by
binary states of 1s or 0s, called bits. All our
emails, images and videos on a computer are
essentially a sequence of these ones and zeros.
Superposition Entanglement
The building blocks of a quantum computer are
called quantum bits, or qubits for short. Unlike
Classical
--------------------------------------------------------------------------- ---------------------------------------------------------------------------
bits, qubits can simultaneously be in a state of 1
and 0 or any probabilistic combination of the two.
The value of classical bits are kept well separated. However, qubits can
The ability to be in two states at the same time is
interact with one another to create entangled states. When an operation
called superposition and is what allows quantum Classical bits are discrete and can is carried out on one qubit, it has an instantaneous impact on other
computers to run a vast number of calculations at only be in one of two states: 1 or 0. qubits it’s entangled with. This is what exponentially increases the
once. 1 0 Repeated calculations on the same
information density in many quantum operations. Referred to by Einstein
as ‘”spooky action at a distance”3, entanglement can happen over any
input of bits will always give the
length of distance.
exact same output, due to their
Classical computers solve problems sequentially deterministic nature.
(one step at a time) and the bits on a transistor
chip are constructed in a way to avoid interference Two unentangled qubits
with one another. However, when the value of one
Quantum
-----------------------------------------------------------------------------
qubit changes, it can affect the value of other Qubits entangled by e.g.
Qubits can have an infinite number operation of a laser. The
qubits regardless of their distance, through a Laser
of values between 1 and 0. These ? ? two qubits now exist in an
process called entanglement. Superposition and superposition values of 1 and 0 can ? ? indeterminate single
entanglement are what allow quantum computers 1 be negative, positive or complex and quantum state
to achieve exponential speedups 1 are represented by a point on a
Bloch sphere2.
? ? Entanglement can happen
Answers given by quantum ? ? over any distance
Qubits can be represented by any two-state 80% computers are probabilistic in
quantum mechanical system that can be nature. When a qubit in
manipulated electronically. The up and down 20% superposition is measured, the vast The measurement of one
states of an electron’s spin or the horizontal and amount of information it carries 1 qubit breaks the
cannot be captured. The quantum entanglement and causes
vertical polarizations of light photons amongst state collapses to a discrete value of the qubit to collapse into
other systems (superconducting circuits, quantum 0 1 or 0 on observation, with an Measurement one of the states 1 or 0.
dots, ions, etc.) can be used to represent the 1s 0 associated probability. This means of one qubit When one qubit is
and 0s that are needed to realize qubits.1 that computations on a quantum measured, the value of the
computer would need to be repeated other qubit is also instantly
many times to converge on the 1 0
0 revealed.
answer with the highest probability.
Source: (1) Accenture Labs, 2017 (2) Boston Consulting Group, 2018 (3) Simonite, 2018
How much faster are quantum

In addition to superposition and entanglement, which
computers?
allow quantum computers to carry out calculations
simultaneously, many quantum algorithms are also
based on the idea of interference. The probability Exponential (significant) speedup – e.g. Shor’s algorithm
amplitudes of different quantum states can interfere to Multiplying two prime numbers (e.g. 3 × 5 ), regardless of their size, is a trivial task. However, it turns out that the reverse of this, finding the prime
factors of a number, is not so trivial. In fact, the difficulty of prime factorisation for very large numbers, is the basis of many common encryption
either strengthen or weaken the probability of solutions
standards, such as RSA, which secure most of our communications over the internet. Classical computers work through problems sequentially, which
by cancelling out paths that lead to wrong answers and means that the complexity of prime factorisation can grow exponentially, the larger the number is. However, quantum computers tackle operations
amplifying those that lead to correct ones. The aim of concurrently, which means the time it takes to find the prime factors of a very large number only grows linearly, as in the case of Shor’s algorithm for
quantum algorithms are to amplify correct answers to prime factorisation. Where it would take a classical computer 300 trillion years to break an RSA-2048 bit encryption key, it would take a 4,099 qubit
quantum computer, only 10 seconds!4 This has significant cyber security implications. As a result of this potential threat, there has been significant
near certainty and settle on a probabilistically correct
research into trying to find new cryptosystems, which are not vulnerable to quantum computers (post-quantum cryptography).
solution.1 For quantum algorithms, unlike classical
ones, an increase in the size of a computational task A near term application of the exponential speedup offered by quantum computers is in the chemicals and pharmaceuticals industry. The difficulty in
doesn’t linearly increase the time required to tackle it. simulating the interactions between molecules grows exponentially as they grow in size (similar to prime factorization).5 Quantum computers could be
used to simulate molecular reactions in a feasible time, which would streamline the process of creating new medicines and materials.
A quantum computer with 𝑛 qubits can conduct
calculations on 2 inputs at once. All these properties
contribute to the extra power and speed-advantages
offered by quantum computing.2 “Where it would take a classical computer 300 trillion years to break an RSA-2048 bit
Two examples of quantum algorithms which offer encryption key, it would take a 4,099 qubit quantum computer, only 10 seconds!”
exponential and quadratic speedups in solving well-
known computational problems, are Shor’s algorithm
and Grover’s algorithm, respectively. Peter Shor Quadratic (moderate) speedup – e.g. Grover’s algorithm
came up with his algorithm for prime factorisation when
The time it takes to search through an unordered list, also increases exponentially with the size of the problem. This is sometimes referred to as the
he was at AT&Ts Bell Labs in 1994. Two years later at ‘phonebook’ problem. If a classical computer had to find someone’s name in a phonebook, based on their number, it would need to go through the list of
Bell Labs, Lov Grover proposed an algorithm for unordered phone numbers one by one, which at maximum would take as many steps as there are entries in the list (the last entry might be the number
searching through unstructured databases.3 The we want). Quantum computing can offer a moderate speedup to this problem. Grover’s algorithm is one of the best known techniques that can offer a
quadratic speed advantage in search algorithms for unstructured databases. Where it would take a classical computer 𝑁 steps to look through a list,
potential speedups offered by these two algorithms
Grover’s algorithm could do this in 𝑁 steps.6 This means, for a list with a billion entries, Grover's algorithm could find the answer in only 31,623 steps!
was one of the major motivations behind efforts to
actually build quantum computers, which at the time However, since the speedup offered is only quadratic in nature (e.g. if a task takes 16 hours using a classical computer it would take roughly 4 hours
were (and still to some extent remain) hypothetical using a quantum one), it might not always justify the expense of using a quantum computer. Still, one area this might be advantageous in, is machine
devices. learning (ML). Currently, GPUs with parallel processing and specialized graphics are used to tackle unstructured search queries. However, a market of
more that $20 billion in ML applications of quantum computing and replacement of GPUs is likely to emerge by 2030.7
Source: (1) Boston Consulting Group, 2018 (2) McKinsey Quarterly, 2020 (3) Deloitte University Press, 2017 (4) QuintessenceLabs, 2019 (5) Boston Consulting Group Henderson Institute, 2018 (6) Deloitte
Insights, 2019 (7) Boston Consulting Group Henderson Institute, 2018
Qubits are extremely sensitive to noise. The slightest

interaction with their environment can cause them to
So why don’t we have quantum
fall out of superposition and lose the information they
were carrying, before they have been fully utilised. The computers yet?
disappearance of a qubit’s quantum properties is
called decoherence.
Decoherence means qubits are highly unstable and

after about 50 microseconds become prone to errors.1
Therefore, controlling qubits and their operations is
very difficult. Error-correcting algorithms and the
Insight
addition of more qubits are ways in which these errors The phase of quantum
can be mitigated. However, around 1000 physical development we are currently in,
qubits are required to create just 1 fault-tolerant and is referred to as NISQ (noisy
error corrected qubit, known as a logical qubit. intermediate-scale quantum). In
this era of development, quantum 50 microseconds - 273 1000 to 3000
Roughly 200 logical qubits and hence 200,000 physical chips will be limited to roughly
qubits are needed for most commercial applications of 50-100 qubits, and although they The period of time qubits can store The kind of temperatures Number of physical
quantum computing.2 The large number of physical will be able to outperform information before they decay qubits need to be kept at qubits required to create
classical computers in certain to maintain their quantum 1 error-corrected logical
qubits needed for error correction, drastically increases
tasks, the noise in their state
the overhead required to carry out meaningful qubit 7
operations (quantum gates) will
calculations. limit the size of quantum circuits
and operations that can be
performed reliably.5 Calculations
Since the slightest vibration or interaction with the will need to take place on a
environment can cause a qubit to collapse out of selection of coherent qubits and
allow for some degree of errors.
superposition and into a classical state, qubits need to The calculations will also need
operate in a vacuum environment, be magnetically to be completed in as few steps
isolated and kept in temperatures near absolute zero.3 as possible, before gate
At absolute zero (0 Kelvins or -273.15 degrees
Celsius), all atoms stop moving, which allows for better
(operation) errors
decoherence set in.
and
< 1 nanotesla 1000s of Kgs $ Billions
control over otherwise volatile qubits. The most The progress made in quantum
The strength of field achieved after The weight of quantum computing The cost of a universal,
common quantum chips in operation today are kept at computing in the next decade is
magnetically shielding qubits - hardware. The large cooling fault-tolerant quantum
unlikely to involve error
around 15 millikelvins (0.015 K), well below the 4 K correction, unless some major qubits can be disrupted by the systems and the many wires and computer
temperature of interstellar space.4 The systems breakthrough is made in this slightest magnetic field. This is apparatus needed to electrically
required for this cooling are extremely costly and can field. Therefore, research and approximately 50,000 time less communicate with individual qubits,
weigh as much as a small car. As a result, it’s very trials are likely to be done using strong than the Earth’s magnetic mean these devices can weigh as
NISQ devices. field 6 much as a car
hard to modulate and scale quantum computers when
such large mainframes are required.
Source: (1) Deloitte University Press, 2017 (2) Deloitte Insights, 2019 (3) McKinsey Global Institute, 2016 (4) Deloitte Insights, 2019 (5) Preskill, 2018 (6) D-Wave Systems, 2016 (7) Boston Consulting Group Henderson Institute, 2018
Applications
What types of problem can a

The answer a quantum computer provides is
probabilistic and making sense of this is a challenge.
quantum computer solve?
Unlike classical computers, quantum computers don’t
offer a single answer. Instead they provide a narrow The types of problems quantum computing is best suited to solve include: optimization, simulation and
range of possibilities and the same calculation can
machine learning.
yield different answers each time it’s carried out.
Therefore, quantum computations need to be repeated
many times before they converge on the most Optimization problems exist in almost every industry and normally include finding the most efficient path or
probable answer. way of solving some problem, which can save costs and increase efficiency. For example, supply chains in
many businesses could benefit from optimisation as there are many variables at play, such as costs,
routes, variation of products, clients and so on.2 The number of options in such multi-variable problems can
For this reason, quantum computing is not the best Optimisation grow rapidly, which is why the speedups offered by quantum computing are well suited to solve these
problems. For example, Grover’s algorithm could be used for optimization problems, where millions of
solution to all problems and will most likely form part of solutions would have to be tested sequentially through a trial and error process otherwise. Optimisation
a hybrid solution with classical computers, rather than examples include, but are not limited to: portfolio optimisation, fault analysis for building stronger planes
entirely replacing them. Where narrowing down the and ships, network design, oil well optimisation, creating new materials, traffic flow optimisation, optimising
answers to a problem or simulating vastly complex battery designs, parking and e-charging search for autonomous vehicles, etc.
systems, proves an essential part of an approach, a
quantum computer is likely to be used. For example, Quantum computers can be used to simulate complex systems and processes such as the interaction
optimising delivery routes for a company can be an between atoms and molecules. The difficulty in simulating such interactions grows exponentially with the
extremely complex task. In this scenario, a quantum size of the problem, meaning it’s impossible for classical computers to accurately represent such systems.
computer could be used to find a subset of the most Richard Feynman, who was one of the early theorists of a quantum computer, said "Nature isn't classical,
Simulation dammit, and if you want to make a simulation of nature, you'd better make it quantum mechanical.“3 In fact,
efficient routes (the most time consuming part of the simulating quantum systems for their use in the field of physics, was one of Feynman’s main motivations
problem), and then a classical computer could give a behind his work on quantum computing. Examples of simulations that quantum computers can generate,
definitive answer in choosing the best route amongst include: simulating enzyme and protein reactions for drug development, quantum chemistry, Monte Carlo
them.1 simulations for risk profiling, traffic simulation, simulating weather systems, etc.
The large size of quantum computing mainframes and Machine learning (ML) is dependent on both sampling and optimisation techniques. Sampling problems
the specialised environments in which qubits need to involve selecting a subset (sample) from a population, which best represents the general characteristics of
be kept means that quantum computers are unlikely to that population. As qubits are inherently probabilistic, it makes it much easier for a quantum computer to
generate random samples.4 In ML, sampling can be used to create better training data, which is what the
ever replace our desktop PCs or mobile phones. Machine Learning ML algorithms essentially learn from. Therefore, the higher the quality of this training set, the more accurate
Additionally, the large costs associated with building and powerful the ML capabilities will be. Many ML algorithms are eventually used to solve optimisation
these devices means that they are only likely to be problems, which as discussed, can be greatly improved by quantum computing. Examples of applications
used when a High Performance Computer cannot include: accelerating the arrival of driverless vehicles, improving fraud detection, better detection of
abnormalities on medical scans, and generally any application ML can be used for.
solve the problem in a reasonable time.
Source: (1) Accenture Labs, 2017 (2) Boston Consulting Group, 2020 (3) Gil, 2016 (4) Accenture Labs, 2017
Quantum computing has the potential to revolutionise

personalised and precision medicines, improve protein
folding predictions, accelerate genomics and streamline Biopharma and life science
the process of drug development. An example of this
power lies in the ability to simulate enzymes. Enzymes
can catalyse a wide variety of biochemical interactions
applications
and target cells very precisely. Harnessing the power of
enzymes and other proteins can lead to new and
personalised drug discoveries. However, to do this, we 31% of life science organizations planned to begin evaluating quantum
need to able to model their molecular structure and computing in 2020 and 39% percent are either planning an evaluation in 2021
interactions. This involves simulating the interaction of
each individual electron and nuclei with one other, as well
Insight or to place quantum computing on their radar.6
as the quantum mechanical effects that occur on the sub- The impact of quantum computing on the biopharma industry is one of the most promising amongst all of quantum
In September of 2017, IBM computing’s potential applications. Since the interaction of molecules with one another is itself a quantum process,
atomic level. This becomes exponentially harder the simulated the largest molecule quantum computers can deliver significant benefits to molecular simulations with relatively low resources. Where it
larger a molecule is. Using a classical computer, it is at the time, Beryllium hydride would take a classical computer as many bits as there are atoms in the universe (1082 bits or one-hundred thousand
practically impossible to simulate structures like enzymes (BeH2), using a quantum quadrillion vigintillion bits) to simulate the 41 atom penicillin molecule, it would only take a quantum computer 286
computer.3 Even though this 3 qubits. We can expect to see hybrid quantum-classical approaches to simulating molecular structures by 2025.7
which are made up of 1000s of molecules. Therefore, to atom molecule was simple
develop new medicines, scientists are forced to actually enough for a classical computer
Hardware/end-to-end providers driving biopharma applications of quantum computing include Google, IBM,
create molecules in the lab using synthetic chemistry. to model, it lay the foundations
Honeywell, D-Wave, Xanadu and Rigetti. Specialist software providers in this field include ProteinQure
They then physically test these molecules to discover for the simulation of more
(collaborating with AstraZeneca),8 ApexQubit, GTN, 1Qbit (in collaboration with Accenture Labs to speedup the
complex structures. In the
their properties and see how they behave. However, August of 2020, Google
discovery of drugs for neurological diseases like Parkinson’s),9 Menten AI (has received funding to design a peptide
these tests don’t always go as planned and the molecules to fight COVID-19),10 Zapata Computing and Qulab. There are also a host of companies that are using high
performed the largest quantum
performing classical computers for ‘quantum inspired’ approaches, alongside machine learning.11 These include
often do not behave as expected, which means more simulation to date using 12
Microsoft, Silicon Therapeutics, XtalPi, Cloud Pharmaceuticals (in collaboration with GSK),12 Atomwise,
trials and tests need to be carried out. Unfortunately, each qubits form its Sycamore chip -
Turbine, and Benevolent AI. All this activity will hopefully speedup and improve the surprisingly inefficient process of
a “Hartree-Fock simulation of
cycle of testing is expensive and lengthy, which is one of diazene isomerization”.4
synthesising molecules. Commenting on the inefficiency of this part of the drug development pipeline, Denis
the primary reasons why developing new drugs takes so Farnosov, the founder and CEO of ApexQubit said, “Just imagine if we plan to invent a new space rocket and needed
to physically build and then launch in order to see how different components work together in the ship”.13
long.1 In the future, quantum
simulations could be used to
model novel coronaviruses and Current inefficiency of the drug development cycle14:
As the interactions of molecules and atoms with each their treatments much faster,
other is itself a quantum process, quantum computers are saving many lives and
well suited to simulating molecules. Quantum computers resources. Kuano, a start-up
The cost of bringing a new 10-15 The time it takes from detecting a
can consider all possible interactions at once and come to founded in 2020, which uses > $2bn new disease to launching a
quantum and AI solutions to medicine to market. years medicine to combat it.
the lowest energy state of a molecule, representing how it screen and develop new drugs,
interacts. Therefore, they would be able to model the has joined the COVID-19 High
structure and interactions of enzymes and many other Performance Computing (HPC)
Consortium to help discover Of drugs that make it to clinical trials,
molecules in just hours, which would drastically shorten drugs that can combat COVID- fail in the first phase of testing. Less
90% Of cancer drugs fail in
the life cycle of drug development and open the door to 19.5 than 10% of all medicines that enter 97% clinical trials.
creating more personalised and targeted therapies. It is clinical trials, make it to market.
estimated that quantum simulations could be worth $20
billion in the pharmaceuticals industry by 2030. 2
Source: (1) McKinsey Quarterly, 2020 (2) Boston Consulting Group Henderson Institute, 2018 (3) IBM News Room, 2017 (4) Science, 2020 (5) Kuano, 2020 (6) QED-C, 2020 (7) Boston Consulting Group,
2019 (8) Businesswire, 2020 (9) Newsroom.accenture.com, 2017 (10) Biopharmatrend.com, 2020 (11) ) Boston Consulting Group, 2019 (12) Biopharmatrend.com, 2020 (13) The Quantum Daily, 2020.
Apexqubit (14) Innovatorsunder35, 2018
Quantum computing (QC) could be used to streamline

many carbon intensive process, contributing to the fight
Quantum computing applications
against climate change. NISQ devices in the next
decade, comprising of about 50-150 qubits will be for tackling climate change
capable of accurately simulating and analysing
molecular reactions before needing to even synthesise a
single molecule.1 This allows for the modelling and To produce ammonia-based fertiliser, the Haber- QC can help find new catalysts for splitting water
Bosch process is used, which consumes 3-5% of molecules to produce hydrogen. Currently,
creation of more efficient chemical catalysts that can all natural gas produced worldwide. However, using expensive platinum and methane (a green-house
improve carbon intensive industrial processes such as bacteria to produce ammonia naturally, takes gas) are used in the process. Using more effective
catalysts in the synthesis of hydrogen, to produce
significantly less energy and QC could be used to
those involved in the production of fertilizer, which accurately model the catalyst (e.g. the FeMoco hydrocarbons, could also lead to the development of
molecule) required to do this. Google’s CEO cheaper emission-free fuel in shipping, aviation and
accounts for 2% of global CO2 emissions!2 predicts that the Haber-Bosch process will be transportation and help solve the storage problem in
obsolete in the next decade.4 renewable power systems.8
Fertiliser production Hydrogen production

After the era of NISQ comes error-corrected quantum
computing which will be capable of solving higher-level QC could help with the design of more efficient, The current ability of classical computers to model fluid
dynamics, used in the design of aircraft, cars and ships
optimisation problems. This can greatly streamline the denser batteries. This could increase the uptake of
is substandard. Drag and lift for aircraft could be better
electric vehicles, which would help reduce CO2
formulation of products and optimise the field of emissions. IBM and Daimler AG (Mercedes-Benz’s optimised using QC, which would reduce emissions.
parent company) are working towards creating next- Airbus is looking at how QC can improve aircraft climb
material science. Materials can be optimised to be generation lithium sulphur (Li-S) batteries that will last trajectory (important in short-haul flights), and how to
stronger, lighter, cheaper and better insulators that longer and be cheaper than today’s batteries. To do improve wingbox design to optimize the weight that
aircrafts can carry. Marine and aviation both account
this they are using QC to model the molecules
require less carbon and resources to be manufactured. required in the production of these Li-S fuel cells.5 for approximately 2% of global emissions9 and both
More efficient batteries industries can benefit from these kinds of
For example, stronger and lighter replacements for Fluid dynamics optimisations.
energy-intensive materials such as aluminium and steel,
can be used in the manufacturing of cars, ships, planes
and buildings. The energy storage density of batteries Binding carbon from the atmosphere and at the Supply chains could also benefit from QC’s
could also be optimised by utilising new materials, which source of its production, helps with reducing the optimisation capabilities as there are often many
variables at play, such as cost, routes, variation of
amount of CO2 in the atmosphere (carbon-
might increase the uptake of renewables (e.g. solar capture). QC could be used to model better products, clients and so on. Improvements in
catalysts which would make this process more supply chain optimisation offered by QC, coupled
power) and electric cars. Additionally, quantum effective and lower its cost. Scalable negative- with driverless cars and more shared economy
computing based ML could be used to optimise yields emission ‘direct-air capture’ solutions could also transportation trends, could significantly reduce
be developed using these catalysts. 6 the carbon footprint of the transportation
and decrease by-products in various industrial industry.10
Carbon capture Supply chain optimisation
processes.
Volkswagen, in partnership with D-Wave, are

Quantum computing can also generate great value and Cement production accounts for 5.5% of CO2
emissions globally. Polymer cements are
using QC to optimise traffic management systems.
This will allow for better traffic forecasts and allow
reduce emissions by improving the design of systems stronger, more resistant to chemicals and have a taxis and public transport to more efficiently deploy
lower carbon footprint than traditional cement.
involving fluid dynamics, such as cars, planes and However, they are only used in a very limited
their services, reducing the waiting time for
customers. This will also avoid taxis and buses
ships. The optimisation of supply chains and logistics number of fields because of the cost of producing from travelling long distances without passengers,
them. QC could give us better solutions and which would help reduce the number of cars on
for planning/routing or product distribution, as well as the formulations for creating this type of cement. 7 the road and thus the volume of CO2 emissions.11
optimisation of traffic systems could also reduce the
Cement production Traffic flow optimisation
carbon footprint of many industries.3
Source: (1) Boston Consulting Group, 2020 (2) Karma, 2020 (3)-(4) Boston Consulting Group, 2020 (5) IBM Research Blog, 2020 (6)-(10) Boston Consulting Group, 2020 (11) Volkswagenag.com, 2018
Financial services applications
Dynamic portfolio Risk profiling and

optimisation aggregation
Quantum computing’s ability to accurately simulate risk
scenarios, optimise portfolios and quickly sift through Portfolio optimization using Monte Carlo simulations can be greatly improved Amendments to regulations such as Basel III and Solvency II mean that money
large unstructured datasets to train ML algorithms for with the arrival of quantum computing.3 Many players are already investing in managers and financial institutions need to simulate a much larger set of risk
fraud detection and enhanced customer targeting, quantum research in this field, including Goldman Sachs, JP Morgan Chase scenarios, which increases the cost of compliance, especially if there are
and Citigroup.4 penalties involved. Quantum computing can help test vast sets of risk-
can deliver significant benefits to the financial services assessment scenarios with more accuracy, driving down the cost of risk profiling
industry. The complexity of financial markets are increasing. This can be attributed to the and making it more efficient. The power of quantum computing could also allow
increasing number of options available for money managers to invest their for real-time risk analysis and financial forecasting.7
money in, alongside the increasing pressures from regulators to be more
Insurance companies in particular can benefit from the transparent and increased market volatility. Therefore, the number of variables The simulation of risk scenarios can also greatly contribute to the process of risk
a money manager needs to consider when optimising their client’s portfolio aggregation for insurers. Additionally, quantum computing’s superior simulating
ability of quantum computers to accurately simulate quickly stack up. Quantum computing can compute portfolios that are optimal capabilities can be used for accurate weather forecasting, which is of
weather systems, to better manage catastrophe and over fewer rebalances compared to classical computing, which increases tremendous value to insurers.8 Both these applications can be utilised in real-time
efficiency and cuts costs.5 Additionally, quantum computing based ML can risk analysis and forecasting.
whether-related losses. The risk aggregation capabilities
greatly enhance pattern recognition and clustering, which will allow for the
of quantum computers can also enhance the aggregation of assets and investors in groups that are seemingly unrelated to
underwriting function of insurers. the naked eye.
In May 2019, Willis Towers Watson announced their Customer targeting and Fraud detection
collaboration with Microsoft to transform their risk product recommendation
management and quantification offerings to insurance
and financial services clients, using quantum Small to medium sized financial institutions often lose customers due to not A report from IBM estimates that “financial institutions are losing between USD
computing.1 targeting them with personalised enough services and relevant products.6 It 10 billion and 40 billion in revenue a year due to fraud and poor data
takes a classical computer an impractically long time to go through large data management practices.” Fraud detection using classical computing is very
sets to find patterns and useful information about customer behaviour. This inaccurate and has an 80% false positive rate, which means financial institutions
means financial institutions (and well as businesses in other industries) are such as banks are overly risk averse. As a result, it often takes banks as long as
Allianz X, the digital investment branch of the Allianz missing the opportunity to better predict customer preferences and target them 12 weeks to onboard new customers due to the process of carrying out credit
group, alongside RBS, have invested in the funding with more pre-emptive product suggestions and personalised services. Quantum checks. In a climate where 70% of banking is done online, banks risk losing
computing’s ability to efficiently and accurately analyse large datasets to find customers with such slow sign-on speeds.9
round for the quantum computing start-up 1QBit.2 patterns and make projections using ML, could help target customers more
effectively, increasing customer retention and satisfaction. The translation of Accurate fraud detection relies on pattern recognition and natural language
classical datasets into quantum ones is one of the main barriers to achieving this. processing (NLP), which is carried out using neural networks (a type of ML
algorithm). In finance and insurance, this can help detect patterns of fraud and
automated attacks. Training neural networks on big data is very difficult using
classical computers. However, quantum computing’s powerful speedups in
searching through unstructured data, can greatly improve this process.
Source: (1) Willis Towers Watson, 2019 (2) Finextra Research, 2017 (3) Atos, 2018 (4) Modelomni, 2020 (5)-(7) IBM Institute for Business Value, 2019 (8) Singularity Hub, 2017 (9) IBM Institute for Business Value, 2019
The quantum
landscape
The current state of quantum computing development is akin to

when classical computers were still using vacuum tubes, before Competing technologies
their transition to silicon transistors. Competing qubit
technologies are in a race to prove themselves the most viable
option for a universal quantum computer – that is a quantum
computer capable of tackling all quantum problems.
Qubit technologies
Error correction remains one of the most resource intensive and

The most popular universal quantum computing architecture today uses superconducting qubits. These are two-level energy systems that use miniature
costly parts of producing a universal quantum computer. loops of superconducting wire, cooled to near absolute zero.5 They are the most studied qubit technology so far and used by industry giants such as
Significant development in this field, could see a workable Google, IBM, Intel, Rigetti and Alibaba. Superconducting qubits benefit from fast gate (operation) times compared to other qubit technologies. However,
quantum computer hit the market as soon as 2028-2030. the challenges facing quantum computers made from these types of qubits are the costs associated with their cryogenic cooling as well as their fast
Microsoft are currently working on a technology that aims to have decoherence times. Their short-lived nature means that for each logical qubit of this kind, 1000s of physical qubits are required for error-correction (the
a 1:1 logical to physical qubit ratio.1 largest superconducting chip today, Google’s Sycamore chip, has only 72 physical qubits). Millions of physical qubits would be required for universal
quantum computing.6
Despite the array of qubit technologies currently in development, A popular alternative to superconducting qubits is to use individual ions trapped in electromagnetic fields (trapped ions). The isolation of individual atoms
the majority of quantum computing end-users will not have to increases their decoherence time, creating more stable and less error-prone qubits. IonQ, Alpine Quantum Technologies and Honeywell are pursuing
worry about committing to a particular qubit implementation when this type of technology. Despite producing higher quality qubits and capable of operating at room temperature, trapped ion qubits are not as popular or
wanting to access quantum computing capabilities. This is developed as their superconducting counterparts. 7 This is because they rely on a less mature technology for production, compared to the already
because most users will access quantum computers via APIs standardised and well known semiconductor technology used for superconducting qubits. Even though trapped ions don’t need to be cryogenically cooled,
they still require an ultra-high vacuum to operate in.8
through the cloud. Cloud offerings allow the high costs involved in
building, running and maintaining quantum computing hardware
to be distributed amongst many users, making it a more There are also a host of other more nascent qubit technologies in development, which if successfully implemented, might prove to be easier to manufacture
accessible and affordable technology across the board. Many at scale compared to superconducting qubits or trapped ions. These include photonic qubits, silicon-based qubits, spin qubits, diamond qubits and neutral
atoms. VCs have been increasingly investing in these alternative hardware solutions since around 2017.9 Another potentially ground-breaking qubit
firms are already offering access to their quantum computing implementation comes in the form of topological qubits, which Microsoft have been researching for several years now. Presently only theoretical,
hardware via the cloud. These services include IBM Q topological qubits based on the exotic Majorana quasiparticle, promise to deliver unparalleled low error rates of 1 part per million (or even per billion)! These
Experience by IBM, Leap by D-Wave, Cirque by Google, Forest qubits could truly change the game and greatly accelerate the arrival of a universal quantum computer.10
by Rigetti and Aliyun by Alibaba.2 Amazon Web Services and
Microsoft Azure have also joined the Quantum Computing as a Last but not least, is an altogether different approach to quantum computing, called quantum annealing, most notably pursued by the Canadian firm D-
Service (QCaaS) space by offering users access to a range of Wave. D-Wave are the first company to build and sell a full-scale quantum-based computer, called a quantum annealer. It is generally accepted that D-
quantum technologies via the cloud.3 Wave’s machine can’t technically be classified as a quantum computer since it is not a general-purpose machine. It’s instead designed to specifically solve
optimisation problems (by finding the global minimum in a given energy landscape). D-Wave’s quantum annealers are based on a special type of
superconducting qubit that also need to be cooled to temperatures near absolute zero.11
Cloud offerings of quantum computing increases accessibility and
familiarity and allows users to investigate potential applications
Number of qubit development projects at universities, government labs and public/private companies (as of Sept 2020) 12
starting today. This will likely increase adoption and demand in
the longer-term. Not having to own the quantum computing
hardware also means that companies won’t have to worry about Superconducting Trapped ion Photonic Spin Cold Diamond Topological Annealing
the VHS vs. Betamax risk when choosing between different qubit
technologies to invest millions in!4 25 20 21 19 10 7 7 7
Source: (1) Boston Consulting Group Henderson Institute, 2018 (2) Boston Consulting Group, 2018 (3) McKinsey Quarterly, 2020 (4) Quantum London, 2020 (5) Nature, 2020 (6) The Quantum Daily, 2020. TQD Exclusive
(7) Boston Consulting Group, 2018 (8) The Quantum Daily, 2020. TQD Exclusive (9) Nature, 2020 (10),(11) Boston Consulting Group, 2018 (12) Quantum Computing Report, 2020. A Tour Through the Quantum Ecosystem
The quantum computing ecosystem

An exciting global quantum computing ecosystem has emerged in the last
decade, off the back of increasing public and private investments into
quantum technologies. More than $700 million has been invested in over 60
different quantum technology companies globally, from 2012 to 2018 (the true
volume of investments is almost certainly higher with many deals kept
secret).1 The trend in private investments has been steadily increasing over
the decade, with firms receiving at least $450 million in 2017 and 2018,
compared to only $104 million in 2015 and 2016. Players involved in
developing quantum hardware have received the majority of VC investment to Key players in different layers of the quantum computing technology
date. Nonetheless, firms specialising in quantum software raised more than A blooming ecosystem stack
$110 million between 2012 and 2018, despite their software not delivering
any real-life benefits as of yet. These firms, which are often selling software Over the recent years, a vibrant ecosystem has
developed around quantum computing technology.
for hardware that has not yet fully matured, are primarily attracting clients in
This ecosystem consists of academic institutions,
fields such as the aerospace industry, which often plans years ahead, the a host of public and private companies, from tech
Software and services End-to-end providers Hardware providers
biopharma industry, which expects to see real applications in the next few giants to start-ups, involved in bringing the providers
years or those in the financial industry, where small competitive advantages hardware and software to market, VCs who want
can translate to huge gains. According to Yianni Gamvros, the head of to invest in this technology and clients who are Zapata Computing Intel
business development at the quantum software company, QC Ware, “This beginning their quantum journey by identifying Quantum Circuits
problems quantum computing can solve and Aliro Technologies
seems like a small investment to get ready for another potentially disruptive IBM
conducting preliminary experiments with the help ColdQuanta
force”.2 QxBranch
of consultants in the field. Google
Bleximo
The key players involved in developing the BosonQ
Historically, North America has dominated the top spot in attracting private Microsoft
investment in quantum technologies. However, the heavy commercialisation technology itself can be broadly separated into IonQ
three categories:
QC Ware
of hardware and software technology in China, which can be seen through Rigetti Computing BraneCell
their intense activity in patent filings, puts this top spot in question. Between AppliedQubit
Hardware providers
2012 and 2017, more than 43% of quantum technology patents originated D-Wave Alpine Quantum
form China.3 Additionally, China has overtaken the US in the number of Providers of quantum computing hardware. This
Cambridge Quantum Technologies
scientific papers published in the field since 2013.4 includes the array of qubits needed for Computing
Honeywell Oxford Ionics
calculations, control systems and the components Avanetix
involved in error correction. Origin Quantum Computing Quantum Circuits
The geographic distribution of private investment in quantum technologies 1QBit
mirrors research hotspots around the world and shows that the US, China, Software and services providers
A*Quantum Alibaba Delft Circuits
UK, Canada and Australia are leading the way. Public funding in quantum
Providers that offer an interface for users to Quantum Computing QuTech
technologies has also been a major driver of its growth over the years. The Quantum Machines
translate classical problems into a quantum Inc Atom Computing
USA, Canada, UK, EU, Netherlands, Germany, China, Russia, South Korea,
computing readable format and vice versa. These Agnostiq
Japan, Australia, Singapore, India, Israel and France have collectively Xanadu
players provide quantum computing functionality kiutra
contributed $22 billion to national public initiatives (as of September 2020).5 through APIs via the cloud. Many software ApexQubit
China leads in public funding with a $10 billion quantum programme, $3 providers also offer consulting services to help Alice&Bob
Amazon
billion of which is dedicated to quantum computing.6 firms identify quantum computing use cases and
Beit QuiX
strategies.
In 2013, the UK government announced the National Quantum Technology End-to-end providers
Programme which included £270m funding for research and outreach. £120m
of this was put into the creation of four “Quantum Hubs” around the country, Hardware providers who also offer cloud-based
opensource software platforms with a varying level For a more comprehensive list of various players across the quantum computing technology stack and
including the National Quantum Information Technologies (NQIT) computing
of access to their hardware.9 more information about each player, visit the Quantum Computing Report website at:
hub, headed by the University of Oxford.7 A further £153m was invested in https://quantumcomputingreport.com/privatestartup/ 10
2019 in the second phase of funding.8
Sources: (1) Boston Consulting Group, 2018 (2) ,(3) Nature, 2020 (4) Atos Ascent Thought Leadership, 2016 (5) Qureca, 2020 (6) Boston Consulting Group, 2018 (7) Atos Ascent Thought Leadership, 2016 (8) Qureca, 2020 (9) Boston
Consulting Group, 2018 (10) Quantum Computing Report, 2020. Private/Startup Companies
For practical applications of quantum computing, we need

processors with enough qubits to run the applications and
The future of quantum development
algorithms that can solve the mathematical problems
behind them. Given a Moore’s Law rate of development in
the number of qubits, without any improvement to error
correction, the quantum applications market is estimated to
reach $2 billion by 2035, and more than $260 billion by IEEE standards for quantum computing nomenclature and benchmarks
2050. With improvements to error correction, this is likely
to rise to $60 billion by 2035 and $295 billion by 2050 (for Whilst the quantum computing industry is growing at a rapid pace, with many different qubit
reference, the global computing market today is worth architectures in development and a growing ecosystem consisting of students, academics,
Insight hardware and software developers, engineers, clients from a broad range of industries and
$800 billion).1
Universal quantum computers are still investors, it lacks a cohesive framework of communication. The nomenclature and benchmarking
at least a decade away. However, standards across the industry are fragmented, making it difficult for various players to
Once a certain level of technical viability is reached and many investors are expecting to see communicate effectively.
the appropriate quantum algorithms to solve industry returns on their capital much sooner
problems are developed, the adoption of quantum than that. Some VCs are hoping that a
computing is likely to take on an s-curve pattern. For technological breakthrough will make In an effort to build a common communication framework, the Institute of Electrical and
a general purpose quantum machine Electronics Engineers (IEEE) announced the IEEE P7130™ — Standard for Quantum Computing
applications where quantum computing can deliver a possible in the next 5-10 years, whilst Definitions project in August 2017. This project aims to standardise nomenclature and
significant speed advantage, there could be 70% adoption others believe that industry applicable terminology across the industry in the hope of reducing confusion and making it a more
in 5 years, similar to the trend of GPU adoption for ML applications will be found for NISQ-era
accessible space to the various players. This is an ongoing project which will update terminology
applications. For moderate speedups, there could be 50% devices in the next few years. There
are also those that are hedging their as progress in made in the field.5
adoption in 15 years.2 bets on the start-up or firm they are
investing in, making enough progress
At the same time, better benchmarks and metrics are needed to measure quantum computing
Executives would be wise to track key developments in that someone else buys them out.
progress. It’s important that these metrics and benchmarks account for the differences in the
quantum computing technologies so that they are ready to However, if these quantum various underpinning technologies. The IEEE P7131™ - Standard for Quantum Computing
put together a quantum team when meaningful investments don’t turn around a profit Performance Metrics & Performance Benchmarking project was launched in 2018 to address this
applications are on the horizon. The indicators of soon enough, there is a danger of a very challenge. This standard aims to define technology agnostic metrics and come up with
development to look out for include: the number of qubits quantum winter dampening the standardised benchmarking across the industry. This will allow different quantum computing
current buzz around this technology. technologies to be benchmarked against each other as well as against classical computers.6
that can coherently be involved in calculations (around 150 Akin to the phenomenon of AI winters,
are needed for quantum simulations), performance Some current benchmarks in use include the quantum volume metric by IBM, implementations of
a quantum winter describes a waning
benchmarks of different algorithms on different qubit interest in quantum computing from
Shor’s and Grover’s algorithms, randomized benchmarking and gate-set tomography.
architectures, the demonstration of quantum advantage investors and the public, following a
failure of the industry to deliver on However, there is also a potential danger in establishing metrics and benchmarks too early.
(when a quantum computer can perform a useful task
their promise of practical applications
which a classical computer is incapable of performing in a Simplistic benchmarks could halt innovation as hardware and software developers become
in the near-future. This would halt the
reasonable time), development of error-resistant progress of the field as the money
wholly occupied with trying to perform highly against a particular metric, that might not even take
technologies such as topological qubits and the successful coming in from investors starts to dry into consideration the limitations or properties of their technology.7 Therefore, careful
up.4 consideration must be given to the field of benchmarking and metrics research, as it will
implementation of error correction in other qubit
ultimately define the progress of quantum technologies.
architectures.3
Source: (1),(2):Boston Consulting Group Henderson Institute, 2018 (3) Boston Consulting Group, 2018 (4) Nature, 2020 (5) Standards.ieee.org, 2017 (6) Standards.ieee.org. 2018 (7) Blume-Kohout, Robin J., Young, Kevin., 2019
Quantum
computing’s threat
to cyber security
Encryption is used to securely transfer information between Demystifying encryption - How do we

two parties. A key is used to mathematically scramble
information into an incoherent format (ciphertext), that hides
the true meaning of the communication from any
secure our communications?
eavesdroppers. Another key can then be used to decrypt the
information into its original format by the intended recipient.
The study of encrypting and decrypting messages is referred Asymmetric cryptography
to as cryptography. The two main ways keys are exchanged Two separate keys are used for encrypting and decrypting: one is private and one is public. Alice wants to send Bob a private message, over an
insecure public channel. To make sure no one can intercept her message to Bob, she needs to encrypt it. To do this, Alice needs Bob to generate a
in cryptography are via public-key (or asymmetric-key) and public key using his own private key. Bob generates a public key and broadcasts it for anyone to access. Alice then uses Bob’s public key to convert
symmetric-key schemes. 1 Insight
Insight her message into an unreadable ciphertext. This then allows Alice to exchange her message with Bob securely, as an eavesdropper wouldn’t be able
to make sense of the message even if they intercepted it. Bob then decrypts Alice’s message using his private key pair, which only he has access to,
and which is mathematically linked to the public key.
When
The we phasetalk aboutof keys quantum in
Public-key cryptography (PKC): A public key available to cryptography,
development we we are
are referring
currentlytoin, Even through the public and private keys are mathematically related, and it is theoretically possible to derive the private key from the public one, the
anyone is used to encrypt information, and a mathematically ais string
referred of data,
to as which
NISQ when (noisy mathematical procedure behind doing so (e.g. finding the prime factors of a very large number) is extremely difficult and technically unfeasible using
a classical computer. However, obtaining the public key using the private key is trivial (e.g. multiplying two prime numbers).6
linked private key pair, which only the intended recipient of the plugged into an accompanying
intermediate-scale quantum). In
message has access to, is used to decrypt the information. algorithm,
this era of can either encrypt
development, quantumor Bob exchanges his public key by broadcasting it
decrypt
chips will some be limitedpiece to roughlyof
This type of encryption allows secure transactions and information.
50-100 qubits,The security they
and although of
communications to take place on open networks such as the encryption
will be systemsable todepends outperform on
internet. Applications include, checking emails, online banking, the mathematical
classical computers complexity
in certainof
editing information on the cloud and the majority of other breaking
tasks, the thesenoise keys.in The their Bob Hello X$%f Hello Alice
exchanges that take place over the internet. Common computational
operations (quantum power required
gates) to will friend Decrypted using M#rt Encrypted using friend
break
limit thecryptographic
size of quantum keys circuits
used Bob’s public key
examples of PKC schemes are RSA and Diffie-Hellman.2 Bob’s private key
today, in a feasible
and operations that time,
can isbe
beyond
performed the capability
reliably. of classical
Calculations
Symmetric-key cryptography: The same key is used to
computers.
will need to However,take place on the a Symmetric cryptography
speedups
selection of offered
coherentby qubits
quantum and
encrypt and decrypt information. The key needs to be computing,
allow for some threaten
degreetoof break errors. The same private key is used for encrypting and decrypting. Again, Alice wants to send Bob a private message, over an insecure public channel. To make sure
no one can intercept her message to Bob, she needs to encrypt it. To do this, Alice and Bob securely exchange a private key with one another. This can be
exchanged securely between the two parties. This can either these
The keys.calculations will also need
done physically (e.g. Bob meets Alice in person ahead of time, to give her a private key) or PKC can be used to exchange the key. Alice then scrambles her
be done physically or using a secure method such as PKC. to be completed in as few steps message using the private key and securely sends it to Bob. Bob can then decrypt the message using the same private key.
PKC,
as symmetric before
possible, cryptography gate
This type of encryption’s safety depends on the safety of and hash functions
(operation) errors are often and The security of this method depends on the security of exchanging the private key. The fact that two users have access to the same key makes it less secure
exchanging the private key.3 used in conjunction
decoherence set in. with one
as it increases the chance of an adversary getting hold of this key. However, symmetric encryption is faster to implement that asymmetric encryption, which
is why it is a preferred method when large amounts of data need to be secured.
another to secure data. For
The progress
example, PKC madecan bein used quantum to
Another important group of cryptographic schemes are hash computingsymmetric
distribute in the nextkeys decade andis Bob securely exchanges his private key with Alice
functions. Hash functions are a one-way process that reduce unlikely
then a hash to function
involvecan error be
correction,
applied, unless the
to upkeep some major
integrity
data to a unique fixed size text (a hash). The smallest change
ofbreakthrough
the data isor made in this
information
to the encrypted data will alter the hash and generate an field. Therefore,
exchanged. research
However, and
these
entirely new hash function. Therefore, hash functions are used trialsschemes
three are likelyare to not
be all
done using
equally Bob Hello X$%f Hello Alice
to maintain the integrity of data and to confirm nothing has NISQ devices. to
vulnerable quantum friend Decrypted using M#rt Encrypted using friend
been tampered with (used in blockchain technology).4 computing.5 the same private the private key
key
Source: (1) Atos Ascent Thought Leadership, 2016 (2) Hudson Institute, 2019 (3) Atos Ascent Thought Leadership, 2016 (4)-(6) RAND Corporation, 2020
The security of most online transactions and data held by

organisations, is based on the intractability of solving the
Quantum computing’s threat to cyber
mathematical problems that underpin cryptosystems.
Examples include the difficulty of prime factorising large security - now and in the future
numbers or solving discrete logarithm problems as in the
case of many PKC protocols such as Rivest-Shamir-Adleman
(RSA), Diffie-Hellman, digital signature algorithms (DSA) Two types of attack
and elliptical curve cryptography (ECC).1 When a breach of
security does occur, it’s normally down to the poor The capability of Shor’s algorithm to break public-key cryptosystems such as RSA and ECC, in just seconds, threatens to undermine today’s
implementation of cybersecurity protocols and frameworks. encryption protocols and put the security of our communications at risk. However, this isn’t a threat that materialises only after universal, fault-
This allows adversaries to steal or manipulate data by finding tolerant quantum computing becomes available. The eventuality of quantum computing means that our communications and data are under
threat, right now! Adversaries are capable of intercepting communication channels today to steal data, which they can later decrypt with the
a way to bypass the encryption systems in place, exploiting
advent of a powerful enough quantum computer. This means, the threat from quantum computing is two fold: cyber attacks both in real-time and
digital or human vulnerabilities instead. retrospectively.
The difference with quantum computing adversaries is that

they will be able to attack the very encryption system itself,
thus adding another layer to the cyber vulnerabilities that – Quantum attacks in real-time
already exist.2 Quantum computing can be used to break One of the biggest cyber threats real-time quantum computing attacks on encryption schemes pose, is the undermining of identity authentication
public-key cryptography (using Shor’s algorithm), and digital signatures. This allows adversaries to pose as anyone and compromise the safety of any network. This type of attack could go
significantly weaken symmetric cryptosystems (using unnoticed for months, allowing many fake banking transactions and wide scale distribution of malware to take place.4 These attacks are only
Grover’s algorithm) and undermine the digital signature possible in real-time, when adversaries have access to quantum computing capabilities (either the hardware itself or via the cloud). Side-channel
schemes used to authenticate blockchain transactions. The attacks also become a possibility with advances in quantum technology. This could for example include, an adversary freezing a stolen credit
lengths of private keys have already needed to be increased card to a temperature where the laws of quantum physics take hold, and then querying information from it in superposition.5
over the years, to make them safe against the continuously It would take an approximately 2,000 qubit, fully fault-tolerant quantum computer to break PKC schemes such as 1,024-bit RSA or 256-bit ECC.
improving power of supercomputers. However, the development of such a device is still more than a decade away.6
Quantum computing doesn’t equally affect all cryptosystems

and can only provide significant speedups in cracking PKC. – Retroactive attacks
Symmetric encryption protocols can be made safe by just
increasing the size of their keys, as quantum computing can
Retroactive quantum attacks are where an adversary harvests data now, and holds onto this encrypted information for about 10+ years, in the
only offer a modest speedup against breaking these hope of being able to decrypt it when a cryptographically relevant quantum computer is available. However, this type of attack only poses a
schemes. For those interested, the type of computational genuine threat to information that needs to be kept secret for a long time. For example, the harvesting of personal credit card numbers don’t
problems quantum computing can easily solve are called really matter as they are likely to expire by the time they are decrypted. However, government and military secrets, Intellectual Property, trade
bounded-error quantum polynomial time (BQP) problems. secrets and some personal videos and messages might need to remain secure for more than 20-30 years.7 Retroactive attacks mean that our
They cannot solve NP problems efficiently but still provide a data is already under threat by quantum computing. Data stolen closer to when fault-tolerant quantum computing is available is the most
valuable and at risk. This is why it is important to take measures now to combat the threat of retroactive quantum attacks. One solution is to
quadratic speedup.3
migrate to encryption schemes that are secure against quantum algorithms (post-quantum cryptography), as soon as possible.
Source: (1) Wallden, 2019 (2) RAND Corporation, 2020 (3) Wallden, 2019 (4) RAND Corporation, 2020 (5) Wallden, 2019 (6) Hudson Institute, 2019 (7) RAND Corporation, 2020
Post-quantum cryptography (PQC) refers to classical algorithms

(normally public-key algorithms) that are secure against quantum
computers. PQC relies on the hardness of mathematical problems
(often in the NP category) that quantum computing algorithms
Post-quantum cryptography
can’t solve in a feasible time. These include hash-based, code-
based, lattice-based, multivariate and symmetric-key
cryptography.1
Standardisation of post-quantum cryptography schemes
Only major developments or an entirely new architecture in fault
tolerant quantum computing, expected to emerge around 2035, There is a need for standardisation in the field of PQC. For a long time,
would pose a threat to current cryptography. However, the threat Insight governments were complacent and did not raise awareness about the threat of
of retroactive attacks and the amount of time it would take to quantum computing to current cryptosystems. Instead, researches interested in
develop, standardise and adopt PQC protocols (most likely As well as PQC, which is a classical solution to quantum-proofing cybersecurity were independently coming up with PQC algorithms and the PQC
communications, there exist other approaches too, such as space was unstandardized and full of untested cryptosystems. Finally, in 2016,
decades), means that governments are looking to implement PQC
Quantum Key Distribution (QKD). QKD is a quantum solution to a the US National Institution for Standards and Technology (NIST) started a
standards much before 2035.2 GCHQ have already started quantum problem. It provides a method for exchanging symmetric
researching post-quantum security3 and the NSA plans to move to project of collating and testing the best post-quantum public-key cryptographic
keys over a public channel, without having to rely on PKC. QKD
protocols. NIST are looking to have a standardised set of PQC schemes
quantum secure standards in the near future.4 Current estimates takes advantage of the properties of quantum systems
themselves to establish security. The very act of observing a published between 2022-2024, after which, a widespread transition to using these
for the number of qubits a quantum computer requires to break new protocols should begin.7 It’s important that the final number of PQC
quantum system will change its state and so any eavesdropper in
PKC schemes, are based on Shor’s algorithm. However, Shor’s the process of key transmission can be detected (and the key algorithms standardised by NIST are kept to an absolute minimum, to avoid a
algorithm hasn’t been shown to be optimal at prime factorisation. discarded). The information in QKD can be encrypted in the fragmentation of approach.
Therefore, a sudden breakthrough in quantum algorithms that polarization and axis of a photon, before it is transmitted. The
sender and receiver can then publicly share part of the encrypted
allow this to be done faster (e.g. using variational quantum To comprehensively test PQC schemes, it is essential to find the fastest quantum
message, and if they differ, they know there has been a breach.
factoring), could bring the timeline of a quantum computer capable QKD signals can be transmitted via optical fibres or wirelessly. algorithms capable of breaking cryptographic keys, and then increase PQC key
of breaking encryption much closer. Consequently, the risk The two primary methods of QKD transmission are via closed lengths in conjunction with this. It’s important that PQC schemes eventually
management approach to quantum computing needs to be QKD secure networks and via nano quantum satellites. adopted are not susceptible to merely moderate speedups offered by more
probabilistic and should consider high impact, low-probability QKD solutions are already commercially available and it is a hot
powerful quantum algorithms discovered in the future. However, most PQC
events as well as longer-term threats. The timeline for threats area of research, including participation from the NSA and GCHQ. schemes suggested so far, such as post-quantum digital signature schemes,
posed by quantum computing should be constantly monitored and The UK Quantum Communications Hub is currently building a often have private/public key or signature sizes that are too large, making them
reassessed. national QKD network which is set to be completed in the next five inefficient to use. In certain fields, where security is the top priority (e.g. defence,
years. China plan to have constructed both a fibre QKD network financial markets, etc.), an inefficient but secure post-quantum cryptosystem is
connecting Beijing to Shanghai, and a quantum satellite. In the not a problem. However, in other industries and particularly for personal use,
Grover’s algorithm threatens to weaken symmetric encryptions quantum race, China are leading in patent submissions for QKD cryptographic inefficiencies (public-key size, key generation speed, encryption
and quantum cryptography, whilst the USA are leading in patents and decryption speed, signature length) are unlikely to be tolerated, and a less
and hash functions only modestly. Therefore, the length of for quantum computing hardware and sensors.6
symmetric keys need to be doubled and the length of hash secure but more efficient protocol would be preferred. Investigation into finding
efficient but secure cryptographic systems is an active field of research.8 There
functions increased by 50% to counteract the threat. Although, Implementing QKD or PQC algorithms is no easy task and might
require large-scale and even national co-operation and are currently no PQC digital signature schemes that offer both short signatures
there also needs to be a consideration for the scalability and and short key sizes.9
standardisation. PQC is less secure than QKD as it only relies on
practicality of decrypting and encrypting using very long keys. the difficulty of solving a mathematical problem, for which
Luckily, there is mathematical evidence (albeit not rigorous proof) increases in computational speed and new algorithms are a
that Grover’s algorithm offers the maximum speedup for a threat. QKD is a safer option but requires the installation of optical
fibre networks and infrastructure. Businesses should decide which
computer to perform a search algorithm.5 This means that PQC
option to choose based on a quantum assessment.
schemes that can withstand Grover’s algorithm, are likely to be
secure against any further quantum developments too.
Source: (1) Wallden, 2019 (2) RAND Corporation, 2020 (3) Atos Ascent Thought Leadership, 2016 (4) Ars Technica, 2015 (5) RAND Corporation, 2020 (6) Atos Ascent Thought Leadership, 2016 (7) Blockchain Research Institute,
2017 (8) Wallden, 2019 (9) Blockchain Research Institute, 2017
Case study: The quantum threat to

The blockchain is a public, decentralised
distributed ledger of transactions (not necessarily blockchain
always financial) in which trust is established
collectively by a peer-to-peer computer network. How the blockchain works
The blockchain consists of blocks of transactions,
each connected to the previous block via some Blockchain transactions have two stages. First is the transaction itself, e.g. sending money to someone, and the second stage is the validation of this transaction. The
linking mechanism, such as the hash of the block validation stage of blockchain transactions consists of nodes in the network validating a group of transactions into a ‘block’. Transactions on the blockchain are time
stamped and digitally signed. To digitally sign a transaction, a user will use the private key within their digital wallet, to create a public key pair which is distributed to
before.
the whole network. The network can then validate the transaction as they know that the public key could have only been generated by the person holding the private
key pair.
The digital signature that authenticates blockchain

transactions uses PKC schemes, which are When majority of nodes agree on the authenticity of a block, it becomes validated and is added to the blockchain (using a hash key). This is usually based on a proof-
of-work (PoW) principle that requires solving some mathematically hard problem, such as inverting a hash function. The first person to validate the block often gets a
vulnerable to quantum computing. The further reward (a new coin in the case of Bitcoin), and so there is an incentive for nodes/users to verify transactions. If an adversary could solve the proof of work of a ‘double
validation and immutability of blockchain data is spend’ (spending the same coin twice), faster than the rest of the nodes in the network, the blockchain would verify this as it would be the longest chain. However,
based on hash functions, which are significantly quantum computing only offers a quadratic speed up to solving proof-of-works and so increasing the key length of hash functions and symmetric keys counteracts this
threat. On the other hand, the digital signature that authenticates blockchain transactions, uses PKC schemes which are vulnerable to quantum computing speedups.
weakened by quantum computers. This makes the forging of transactions and the misdirecting of funds possible. Other security mechanisms of some blockchains, such as a variant of zero-knowledge
proofs called zk-SNARKs, which act to provide anonymity to the users, are also not quantum-safe. This means, years worth of all blockchain data could become
suddenly deanonymized with the advent of fault-tolerant quantum computing.2
A possible future scenario that could unfold when
universal fault-tolerant quantum computing Quantum-proofing the blockchain
becomes available, is that companies who have not
quantum-proofed their blockchain, will face The most important security aspects of blockchain relate to its immutability and the impossibility of double spending. There are two approaches that can be taken to
reputational and financial damage. An adversary quantum-proofing blockchains. Either only future blocks that are added can be made quantum-secure, or all previous and future blocks could be secured (which would
capable of breaking 2048-bit RSA numbers or ECC be useful for existing chains such as Bitcoin). The former is much easier and more cost-effective. The RSA and EC-DSA digital signatures would just need to be traded
for PQC algorithms. The downside of many of these algorithms is the often large increase in key size and computational time needed to implement them. EC-DSA, the
encryption, using Shor’s algorithm, will be able to scheme used in the Bitcoin blockchain today, is 71 bytes on average. In contrast, PQC digital signature schemes are at least 6 times larger. The most promising PQC
forge digital signatures on the blockchain and algorithms for blockchain are Quantum Random Number Generators (QRNGs). QKD as well as other quantum-based technologies, that won’t be available for years
misdirect funds to chosen accounts. Additionally, to come, might also help in quantum-proofing the blockchain. These include quantum authentication, quantum money, and quantum fingerprints.
Grover’s algorithm which can weaken hash
functions, will allow adversaries to rewrite historic Nevertheless, even in a future where the blockchain is quantum-proofed, an asymmetry in the number of users who have access to quantum computing, could still
pose a threat to the PoW principle and make double spending possible. This can happen if quantum adversaries (who in this scenario are the only ones with access to
records and create fake transactions. quantum computers) are able to validate transactions on average faster than everyone else, and so can double spend money without anyone noticing. The PoW for
bitcoin currently takes about 10 minutes and the difficulty of it is increased as time goes by, since more powerful computers join the network. A fault-resistant quantum
computer using Grover’s algorithm can solve this PoW in about one minute. This means an adversary with a quantum computer can double spend and rewrite the
Institutions that use blockchain as a ledger to blockchain history by creating a fork that grows much faster than the original chain, becoming the longer of the two over time and being validating by the rest of the
validate financial transactions, IoT devices that use network. This can essentially erase all previous transactions. Even if those with access to quantum computing decided not to double spend, they would still have a
blockchain for micropayments and the recording of better chance at solving PoWs than others, which could lead to a concentration of power. Therefore, search functions shouldn’t form the solution to solving blockchain-
safe PoWs, as Grover’s algorithm provides a speed up for this type of problem.
state health data, will all be compromised in this
future scenario.1
Making blockchains that are already in existence quantum-secure, requires no one to use a public key twice or for wallets that have already done so to transfer their
funds to another wallet with PQC encryption. This is because a quantum computer can use the already revealed public key to find the private key and manipulate the
funds in that wallet. Implementing PQC for historic blocks can be logistically difficult to do.3
Source: (1)-(3) Blockchain Research Institute, 2017
In assessing whether we should be worried about the threat of

quantum computing to cyber security, we need to ask three key
Should we be worried?
questions:
1. How soon are quantum computers capable of breaking

current encryption schemes likely to be developed? The barriers to PQC adoption and the need for crypto-agility
2. How quickly are PQC algorithms likely to be standardised?
Even though quantum computers with thousands of qubits capable of breaking current encryption protocols are about 15 years away, businesses and governments
3. How quickly is PQC likely to be adopted? need to start thinking about quantum security now. This is because quantum attacks can happen retroactively, and it takes time to create, standardise and
implement, safe solutions. NIST has called for two rounds of PQC submissions to consider in its final list of standardised schemes. Most institutions and other
standardisation agencies, are in anticipation of the results of NIST, which are likely to come out between 2022 and 2024. After the standardised list is published by
The relationship between the above three timelines determines the NIST, the various algorithms suggested will go through years of rigorous testing by the cryptography community, to make sure they are safe. Executives should then
threat level posed by quantum computing. The RAND Corporation be on the look out to decide when it is the right time to start investing in some of these new protocols. The testing phase of the PQC schemes can take years. If and
have attempted to shed some light on these questions, by enlisting when a flaw is found, all the effort put into it developing that solution becomes obsolete. A quantum attack was found against a lattice-based scheme developed in
2007, called Soliloquy, by its creators, which meant this cryptogram had to be abandoned.
expert elicitation. They expect, quantum computers capable of
breaking cryptographic protocols to be available between 2033-
Implementation of PQC or any new cryptographic standard can be a very expensive and lengthy process. Therefore, many firms that conduct a cost-benefit analysis,
2035. This means quantum computing is not the end of encryption in decide that using outdated cryptosystems is preferable to spending the money on new software and hardware. For example, NIST standardised SHA-2, a new
the near future. Although, some experts believe that both much standard for cryptographic hash functions, in 2002. However, 35% of websites were found to still be using certificates with an older standard as late as November
earlier and later developments are possible. A very small minority 2016. Many businesses, particularly those in the financial services industry, have a reliance on legacy hardware/software and an inability to quickly implement new
don’t believe a cryptographically relevant quantum computer will encryption standards. Additionally, companies like to wait until their hardware has served its time before changing it out. It’s thought that the perception of “high
switching costs” and the lack of urgency felt by many companies, means that adoption of new PQC standards would only take place when absolutely necessary and
ever be created, whilst others think major developments mean we when all the old devices and software become obsolete. This means, new, non-PQC hardware installed closer to the time of fault-tolerant quantum computing, could
could have such a device as soon as 2023. The variation in create some serious vulnerabilities. IoT devices, aircrafts and vehicles might be the slowest to adopt PQC, as they are examples of long-lived products. It is expected
predictions about when fault-tolerant quantum computing will be that the adoption of PQC standards might take around 25 years!
possible, means a flexible approach should be taken to managing
the risks posed by this technology. Another barrier to successful PQC implementation is that many large organisations don’t have an adequate inventory of their PKI (public-key infrastructure - a
framework for linking individuals to public keys for digital authentication), and all the potentially vulnerable nodes in their system (especially where there are third
party agreements with other vendors). For PQC implementation to be successful, all nodes of a network need to be protected, which can be a mammoth task. Other
Despite this, RAND assess the threat of quantum computing to barriers to implementing PQC include, hardware and software having set key or signature lengths that are incompatible with PQC, and the general inflexibility of
our security systems to be imminent, considering the time it takes some hardware to handle other encryption schemes. The difficulty in transitioning to new cryptographic systems, including the cost of changing software and
hardware, has spurred on the call for what is called cryptographic agility – the ability to implement new cryptosystems quickly, without the need to significantly
to standardise and implement solutions. Experts believe that an change the existing hardware and infrastructure in place.
almost complete adoption of PQC for technology businesses and
governments in the UK and US (i.e. adopted by more than 95% of
It is likely that the standardisation of PQC precedes the availability of cryptographically relevant quantum computing by almost a decade. In this scenario, even if
organisations) will take place in the mid 2030s. If the implementation most players adopt PQC standards before the threat from quantum computing materialises, there will still be a minority who don’t (there is bound to be some old
of PQC doesn’t take place before encryption-breaking quantum equipment or processes that get forgotten and neglected in the transition to PQC). This could cause problems as it increases the surface area of attack against those
computers are available, this could be devastating.1 players which share a network with these vulnerable nodes. A cyber security framework published by NIST in 2018, recommends that organisations take into
account supply chain risk management (managing PKC risks in the whole supply chain including third party vendors). Companies should consider transitioning to a
non-PKC encryption system where the risk to their whole network is compromised by an unagile node.
PQC schemes and government led approaches are already
underway in countries like the USA and the UK, to ensure that NIST plan to make their efforts and standardisations global and so will need to be in contact with the International Organisation for Standardisation (ISO) to make
quantum computing doesn’t turn out to be the disaster it is capable this happen. However, the global effort to transition to PQC standards is expected to take decades, which is potentially far longer than the time available for the task.
This is why thinking about quantum security now is so important. The process of transitioning to PQC may even be an opportunity to become more cryptographically
of becoming. agile in general, and could lead to a safer security system, regardless of whether the risk from quantum computing materialises or not. 2
Source: (1), (2) RAND Corporation, 2020

Insurance impacts
– navigating the
quantum realm
There are many insurance lines of business that will be impacted by the
emergence of quantum computing. However, the largest potential impact
Insurance lines of business affected
arises from the cyber security risks posed by cryptographically relevant
quantum computers. Cyber risks by their nature can have an influence
on many lines of insurance business and practically all industries. In a
scenario where the cryptographic threat from quantum computing Systemic cyber risk
precedes the full adoption of PQC or other quantum-secure protocols,
the world and the insurance industry face a systemic cyber risk that can
affect everything from the security of our text messages to state guarded Sate sponsored cyber attacks using a quantum computer could be used to break RSA and forge digital signatures. This would allow
adversaries access to private and public networks where they could spread malware to dismantle critical infrastructure. Falsified information
secrets and access to military codes.
could be spread using the accounts of high profile figures. Unsolicited access to weaponry and nuclear codes could be used to wreak havoc.
Classified data held by the military could be decrypted and all operations, whether on land, in sea or in space would be vulnerable as global
Quantum computing also exacerbates the risk faced by AI technology, networks would be compromised (systemic political risk impact).
since quantum computing will improve ML capabilities, leading to a
broader adoption of ML and AI solutions and products in all industries. Other nefarious adversaries, including fraudulent employees (fidelity risk), could access or alter personal, legal, operational or financial data.
The impacts of AI on insurance are further discussed in the Lloyd’s report The PKI, used to distribute private keys and digital authentication certificates in military agencies and many large organisations including
Taking control: AI and insurance.1 The following are few examples of financial institutions, would come under attack. This would allow the forgery of common assess cards (CAC) which are required to access
insurance lines of business that can be impacted by quantum computing: classified networks and data.2 Trade secrets and IP could be stolen (some nation states already use IP attacks as part of their economic
strategy). The security behind robots and IoT devices, which are likely to be employed on a wider scale due to AI improvements offered by
- Product liability and product recall: Liability arises from AI-based machinery quantum computing, would be compromised. This would allow large scale supply chain disruptions ((contingent) business interruption).
and products making a mistake. Whilst the risk of AI malfunction increases Autonomous vehicles could be hacked to divert their path and cause accidents (third-party motor liability). 3D printers and manufacturing
once more AI reliant products enter the market as a result of improved ML machinery connected to the internet could be tampered with, leading to large scale product liability and product recall claims. The blockchain
capabilities, quantum computing will also most likely improve the accuracy of technology underpinning cryptocurrencies could be manipulated to alter or forge transactions and double spend money. The privacy of civilian
AI as larger data sets can be processed to train ML algorithms. Therefore, the messages, photos and medical data would also be compromised. Insurance institutions might be particularly targeted as they hold vast
product liability and product recall risk landscape of robots and AI products will amounts of sensitive policyholder data, resulting in hefty GDPR fines.
be changed with the emergence of quantum computing.
As a result, quantum computing could pose one of the largest scale systemic risks in history. With the ever increasing interconnectedness of
- Third-party motor liability: The liability complications arising from accidents systems and our reliance on digital communications, a technology capable of breaking the very encryption system behind our cyber security
involving autonomous vehicles will become a reality sooner than expected, protocols would have implications that affect every aspect of our lives, from state security, to the privacy of our text messages.
with ML based quantum computing accelerating the arrival of driverless cars.
- Political risks: The power of quantum computing could lead to the creation of
better deep fakes, be used to better distribute online propaganda and fake
news and take advantage of human behaviour for social engineering. This
Information at Private messages
Cloud computing Credit card details cryptocurrencies Intellectual property
increased capability to instigate political unrest can lead to more protests, risk from e-commerce Classified access codes
and accounts
cryptographically to nuclear plants and weaponry
followed by government backlash, which would have an impact on business
interruption and property damage. relevant
quantum
computing
- Property damage: The expensive hardware and control systems employed in Trade secrets Juvenile criminal records State secrets and emails Online banking Genetic information
Computer aided designs
quantum computing mainframes will increase the property risk profile. communications
(CAD) for 3D printing and medical history
Source: (1) Lloyd’s, 2019 (2) Lindsay, 2020

Business opportunities for insurers
“The quantum computing hardware and software market is growing fast, and
“The quantum threat to cybersecurity is an example of
a self-denying prophesy: the more credible the threat
could be worth more than $50 billion by 2030”
narrative, the more concerted the effort to counter it”.1
The cyber threat posed by quantum computing has
spurred an international effort to quantum-secure
Business opportunities
encryption protocols. This provides insurers the The quantum threat to cybersecurity has incited a concentrated effort from governments and research institutions to make sure this threat never
opportunity to offer risk management services to many materialises. Therefore, the systemic cyber threat posed by quantum computing, acts as a catalyst for improving cyber security standards and
business. Insurers can help promote crypto-agility increasing crypto-agility across all industries. This provides the insurance sector with the following opportunities:
within organisations, in an effort to mitigate the cyber - Opportunity to work directly with various clients to provide risk management services. This includes identifying vulnerable nodes in the cyber
risk of quantum computing. They can also promote the security infrastructure of clients and helping them build an inventory of PKC protocols that need to be eventually replaced with PQC schemes.
education of quantum threats within organisations to This would help industries address the quantum threat to cybersecurity before the arrival of cryptographically relevant quantum computing.
better ready risk managers for a timely transition to
- By leading in this space, insurers will be able to gain the expertise and knowledge required to guide their policyholders to a more resilient,
PQC protocols and quantum-agile cyber security cryptographically-agile future, thereby thwarting the potential for a systemic cyber threat in the ecosystems they operate within. By avoiding a
frameworks. systemic cyber catastrophe, insurers also ensure that they will be able to continue providing insurance products to policyholders.
Even though quantum computers capable of breaking

current encryption schemes are still about 15 years The quantum computing hardware and software market is growing fast, and could be worth more than $50 billion by 2030.2 This is a substantial
emerging market with new and unknown risks, that offers insures the opportunity to provide innovative insurance products and services to quantum
away, we can expect to see industry applications of technology developers, distributors and adopters. The development of Quantum Computing as a Service (QCaaS) via the cloud, will also expand the
quantum computing emerge in the next 5 years. This opportunity for SMEs to adopt quantum computing solutions. This growth offers insurers the opportunity to:
provides insurers with opportunities to offer bespoke
services and products to the emerging array of - provide a new host of quantum hardware/software start-ups and companies with insurance to meet their needs. For example, companies offering
quantum applications such as algorithms for enhancing risk modelling, weather prediction or fraud detection, might wish to ensure against their
companies involved in the development, deployment
algorithms making incorrect decisions and affecting clients, by purchasing specialised professional indemnity or cyber products. Hardware
and adoption of quantum hardware and software. developers may wish to purchase property insurance against damage to their expensive equipment.
- provide insurance and create new lines of business off the back of industries that will be spearheaded with the arrival of quantum computing (e.g.
autonomous vehicles, electric cars, renewable energies, material science, patient specific drug development and therapies).
- provide innovative products and solutions to developers and adopters of QCaaS via the cloud, which is likely to be the most common way in
which most companies access quantum computers. Many start-ups that have come through the Lloyd’s Lab offering innovative solutions to
emerging risks. Parametrix, a Cohort 4 start-up, offer insurance for external service downtime such as cloud outages, helping to close the
protection gap in business interruption. The emergence of the QCaaS sector would offer business opportunities for such Insurtechs.
Source: (1) Lindsay, 2020 (2) Boston Consulting Group Henderson Institute, 2018
Operational benefits to insurers
Operational benefits to insurers:

The ground-breaking computational power
offered by quantum computing can deliver
significant operational benefits to insurers. - The ever increasing deployment of IoT devices and sensors in various environments, has led to a boom in the volume and
Munich Re transferred quantum computing from veracity of data available to insurers. Quantum computing is well placed to process this large amount of data, which
the HOLD-level (placed there in 2016 and 2017), enables a much greater understanding of risk and offers opportunities for improved pricing and risk models within the
to the ASSESS stage in 2018 (where it still underwriting process. There are also opportunities to collaborate with clients to share risk relevant data for creating better
remains). products.
- The ability to accurately simulate weather systems delivers significant improvements to catastrophe modelling (used in
property insurance), benefiting the process of pricing, reserving and setting policy limits. The modelling of other
20%
aggregate risk such as supply chain interruption, liability risks or cyber, could also benefit from quantum computing
capabilities
- Improved customer relationship management (CRM). Quantum computing can more accurately target customers and
predict their preferences based on customer behaviour data. This can enhance customer satisfaction and retention by
. targeting policyholders with more pre-emptive insurance products and service recommendations .
of organizations will be budgeting for quantum - Improved natural language processing (NLP) capabilities could lead to better fraud detection, market insights, trend
computing applications by 2023, compared to analysis and predictive analytics models.
less than 1% in 2018.1 Nature’s analysis of
LinkedIn data showed that “21 banks and - Automation of the claims function in real-time using rapid data flow from smart devices, which reduces costs and drives
insurance companies in the US and Europe have efficiency
hired more than 115 people with quantum
expertise as of June 2020.”2 Insurers who are proactive in investing in quantum solutions as they start to emerge, can gain a significant competitive edge
based on the variety of operational benefits quantum computing offers to insurers. Additionally, the accessibility of QCaaS
via the cloud is likely to produce Killer Apps for the insurance sector, which increases the opportunity costs associated with
insurers not investing in quantum solutions.
Source: (1) Munich Re, 2019 (2) Nature, 2020

Moving forward
Moving forward – six actions you could take

There are six actions insurers can take to increase their preparedness in dealing with the potential cyber risks from cryptographically relevant quantum computing and the
opportunity costs involved in not adopting quantum solutions and Killer Apps fast enough. Developing a quantum strategy today will help insurers become cryptographically-
agile, and well placed to start building quantum solutions as soon as the opportunity arises. Quantum computing is likely to have a significant impact on the insurance industry
between 2025 and 2035.
Include the risk from quantum computing in their organisational risk assessment from 2020 onwards, Identify which quantum applications would best suit your future needs 2
and develop five year roadmaps of how to plan for the impacts of quantum computing 1
1 4
- Assess the potential opportunity cost of not adopting quantum computing applications
- Educate risk managers and executives on quantum computing applications and threats, to better help - Decide which quantum computing solution and provider best suits your needs (insurers will not need to
them manage risks invest in multimillion dollar quantum computing hardware. QCaaS will offer organisations the
opportunity to rent quantum computing capabilities via the cloud, for specific tasks they need).
- Keep up to date with latest quantum computing and PQC developments
Source: (1) Novarica, 2019
Implement PQC protocols as soon as rigorous testing of the algorithms has taken place Identify staffing needs 3
2 5
- keep up to date with PQC standards - Quantum talent is not readily available and is unlikely to be able to service the inevitable surge in
- Following the standardisation of PQC algorithms by NIST (circa 2022-2024), insurers might want to demand. Therefore, it’s a good idea to either start thinking about hiring and building a quantum team in
consider implementing PQC protocols, as soon as they have been through rigorous testing. Waiting the near-future, or upskilling existing technically-skilled employees.
to implement these protocols after they have been sufficiently tested, protects against potential - Building a quantum team early on means you will be able to take advantage of quantum applications as
flaws that might exist in the algorithms, which could hinder the organisation in becoming secure soon as they become available, providing you with a competitive edge.
against quantum attacks. .
- Keep an inventory of all the places within the company that public keys are used, as these
schemes will have to be migrated to PQC solutions in the future.
Make sure that communication about the cyber threat of quantum computing to decision-makers in
Develop crypto-agility
your organisation as well as to policyholders, finds a balance between exaggeration and complete
3 6
dismissal.
- Become agile in seamlessly transitioning between cryptographic protocols to best prepare for a swift
transition to the safest PQC schemes, when available.
- The uncertain timeline for the development of a cryptographically relevant quantum computer, makes it
- Modernise all existing protocols in PKI, in a way that permits easy transition to PQC protocols easy to dismiss the threat it poses altogether. However, not taking this threat seriously enough could
have devastating effects on the whole of society.
- All that will happen if the cyber threat from quantum computing does not materialise, is that we will have
ended up with more cryptographically-agile security frameworks that will help to protect us against future
cyber attacks.
Source: (1) – (3) Novarica, 2019

Contacts
Acknowledgements
About Lloyd’s
The following people were consulted or commented on earlier drafts of
Lloyd’s contacts Lloyd's is the world's specialist insurance and reinsurance
the report; we would like to thank them all for their contributions:
market. Under our globally trusted name, we act as the
Dr. Trevor Maynard market's custodian. Backed by diverse global capital and
Head of Innovation excellent financial ratings, Lloyd's works with a global
Members of the Quantum London group network to grow the insured world –building resilience of
Operations
– With special thanks to Paolo Cuomo & Emanuele Colonnella local communities and strengthening global economic
Lloyd’s of London growth.
T: +44 (0) 20 7327 6141
E: Trevor.Maynard@Lloyds.com With expertise earned over centuries, Lloyd's is the
foundation of the insurance industry and the future of it.
Led by expert underwriters and brokers who cover more
than 200 territories, the Lloyd’s market develops the
Author essential, complex and critical insurance needed to
underwrite human progress.
Anahita Zardoshti
Insurance Graduate About Quantum London
Innovation Team
Quantum London are a group of professionals from different
Lloyd's of London industries who are interested in understanding the business
T: +44 (0) 207 327 5829 implications of quantum computing. This community was set
E: Anahita.Zardoshti@Lloyds.com up to help ‘non-technical’ business professionals
understand what quantum technology might mean for their
industry, and over what timelines it might have an impact.
Bibliography
Accenture Labs, 2017. Think Beyond Ones and Zeros. [online] Available at: https://www.accenture.com/t20170613t220433__w__/us-en/_acnmedia/pdf-54/accenture-807510-quantum-computing-rgb-v02.pdf
Ars Technica, 2015. NSA Preps Quantum-Resistant Algorithms To Head Off Crypto-Apocalypse. [online]. Available at: https://arstechnica.com/information-technology/2015/08/nsa-preps-quantum-resistant-algorithms-to-head-off-crypto-apocolypse/
Atos Ascent Thought Leadership, 2016. Quantum finance opportunities: security and computation. [online] Available at: https://atos.net/wp-content/uploads/2017/02/Ascent_White-Paper_Quantum-Finance-FINAL-Nov2016-1.pdf
Atos, 2018. Quantum computing in financial services. [online] Available at: https://atos.net/wp-content/uploads/2018/05/Atos-Quantum-FS-white-paper.pdf
Arute, F., Arya, K., Babbush, R. et al., 2019. Quantum supremacy using a programmable superconducting processor. Nature 574, 505–510. https://doi.org/10.1038/s41586-019-1666-5
Biopharmatrend.com, 2020. 18 Startups Using Quantum Theory To Accelerate Drug Discovery. [online] Available at: https://www.biopharmatrend.com/post/99-8-startups-applying-quantum-calculations-for-drug-discovery
Blockchain Research Institute, 2017. Quantum-Proofing the Blockchain. [online] Available at: https://evolutionq.com/quantum-safe-publications/mosca_quantum-proofing-the-blockchain_blockchain-research-institute.pdf
Blume-Kohout, Robin J., Young, Kevin., 2019. Metrics and Benchmarks for Quantum Processors: State of Play. Sandia National Lab, United States. https://www.osti.gov/servlets/purl/1493362
Boston Consulting Group, 2018. The Next Decade in Quantum Computing - And How to Play. [online] Available at: https://image-src.bcg.com/Images/BCG-The-Next-Decade-in-Quantum-Computing-Nov-2018-21-R_tcm9-207859.pdf
Boston Consulting Group, 2019. Will Quantum Computing Transform Biopharma R&D?. [online] Available at: https://www.bcg.com/publications/2019/quantum-computing-transform-biopharma-research-development
Boston Consulting Group, 2020. A Quantum Advantage In Fighting Climate Change. [online] Available at: https://www.bcg.com/publications/2020/quantum-advantage-fighting-climate-change
Boston Consulting Group Henderson Institute, 2018. The Coming Quantum Leap in Computing. [online] Available at: https://www.bcg.com/publications/2018/coming-quantum-leap-computing
Businesswire, 2020. Proteinqure Collaborates With Astrazeneca To Design Novel Peptide Therapeutics. [online] Available at: https://www.businesswire.com/news/home/20200709005086/en/ProteinQure-Collaborates-with-AstraZeneca-to-Design-Novel-Peptide-Therapeutics
Chojecki, P., 2019. Quantum Computing For Business. [online] Available at: https://towardsdatascience.com/quantum-computing-for-business-347b95d400f9
Gil, D., 2016. IBM Brandvoice: With The Dawn Of Quantum Computing, Let's Build A Quantum Community. [online] Forbes. Available at: https://www.forbes.com/sites/ibm/2016/05/04/with-the-dawn-of-quantum-computing-lets-build-a-quantum-community/#224205af3815
Deloitte Insights, 2019. Technology, Media and Communications Predictions 2019 - Quantum computers: The next supercomputers, but not the next laptops. [online] Available at: https://www2.deloitte.com/content/dam/Deloitte/ua/Documents/technology-media-telecommunications/DI_TMT-
predictions_2019.pdf
Deloitte University Press, 2017. Signals for Strategists - From fantasy to reality: Quantum computing is coming to the marketplace. [online] Available at: https://www2.deloitte.com/content/dam/insights/us/articles/3456_From-fantasy-to-reality/DUP_From-fantasy-to-reality.pdf
D-Wave Systems, 2016. The D-Wave 2X™ Quantum Computer Technology Overview. [online] Available at: https://www.dwavesys.com/sites/default/files/D-Wave%202X%20Tech%20Collateral_1016F_0.pdf
Finextra Research, 2017. RBS And Allianz Join C$45M Funding Round For Quantum Computing Startup 1Qbit. [online] Available at: https://www.finextra.com/newsarticle/31395/rbs-and-allianz-join-c45m-funding-round-for-quantum-computing-startup-1qbit
Hudson Institute, 2019. The Executive’s Guide to Quantum Computing and Quantum-secure Cybersecurity. [online] Available at: https://www.hudson.org/research/14930-the-executive-s-guide-to-quantum-computing-and-quantum-secure-cybersecurity
IBM Institute for Business Value, 2019. Expert Insights: Exploring quantum computing use cases for financial services. [online] Available at: https://www.ibm.com/thought-leadership/institute-business-value/report/exploring-quantum-financial#
IBM News Room, 2017. IBM Pioneers New Approach To Simulate Chemistry With Quantum Computing. [online] Available at: https://newsroom.ibm.com/2017-09-13-IBM-Pioneers-New-Approach-to-Simulate-Chemistry-with-Quantum-Computing
IBM Research Blog, 2020. IBM and Daimler use quantum computer to develop next-gen batteries. [online] Available at: https://www.ibm.com/blogs/research/2020/01/next-gen-lithium-sulfur-batteries/
Innovatorsunder35, 2018. Noor Shaker. [online] Available at: https://www.innovatorsunder35.com/the-list/noor-shaker/?platform=hootsuite
Lindsay, J.R., 2020. Surviving the Quantum Cryptocalypse. Strategic Studies Quarterly 14, no. 2, p. 49–73.
Karma, 2020. Quantum Computing To Be Harnessed In Climate Control Effort. [online] Available at: https://karmaimpact.com/quantum-computers-harnessed-to-solve-climate-change-may-be-on-the-horizon
Kuano, 2020. Kuano Joins The COVID-19 High Performance Computing Consortium. [online] Available at: https://www.kuano.ai/news/blog-post-title-three-jnm9r
Lloyd’s, 2019. Taking control: Artificial intelligence and insurance. Available at: https://www.lloyds.com/news-and-risk-insight/risk-reports/library/technology/taking-control
McKinsey Global Institute, 2016. The growing potential of quantum computing. [online] Available at: https://www.mckinsey.com/industries/technology-media-and-telecommunications/our-insights/the-growing-potential-of-quantum-computing
McKinsey Quarterly, 2020. A game plan for quantum computing. [online] Available at: https://www.mckinsey.com/business-functions/mckinsey-digital/our-insights/a-game-plan-for-quantum-computing
Modelomni, 2020. Are Wall Street Banks Behind In The Technological Race?. [online] Available at: https://modelomni.com/wall-street-banks-bet-on-quantum-computing/#
Munich Re, 2019. Tech Trend Radar 2019 – Top 10 Trends. [online] Available at: https://www.munichre.com/content/dam/munichre/global/content-pieces/documents/FIN_Top10_Tech_Trend_Radar_2019.pdf
Nature, 2019. Hello Quantum World! Google Publishes Landmark Quantum Supremacy Claim. [online] Available at: https://www.nature.com/articles/d41586-019-03213-z
Nature, 2020. Quantum Gold Rush: The Private Funding Pouring Into Quantum Start-Ups. [online] Available at: https://www.nature.com/articles/d41586-019-02935-4
Newsroom.accenture.com, 2017. Accenture Labs And 1Qbit Work With Biogen To Apply Quantum Computing To Accelerate Drug Discovery. [online] Available at: https://newsroom.accenture.com/news/accenture-labs-and-1qbit-work-with-biogen-to-apply-quantum-computing-to-accelerate-drug-discovery.htm
Novarica, 2019. Quantum Computing And Insurance: Overview And Potential Players - Novarica. [online] Available at: https://novarica.com/quantum-computing-and-insurance-overview-and-potential-players/
Preskill, J., 2018. Quantum Computing in the NISQ era and beyond. Quantum, 2, p.79.
Quantum Computing Report, 2020. A Tour Through the Quantum Ecosystem. [online] Available at: https://80b.bf1.myftpupload.com/wp-content/uploads/2020/09/A-Tour-Through-the-Quantum-Ecosystem-September-10-2020.pdf
Quantum Computing Report, 2020. Private/Startup Companies - Quantum Computing Report. [online] Available at: https://quantumcomputingreport.com/privatestartup/
Quantum London, 2020. Building a Quantum Computing Ecosystem with Sergio Gago. [Webinar]. Available at: https://www.youtube.com/watch?v=27l2JBQkunw&feature=emb_logo
QED-C, 2020. Almost One Third Of Life Science Companies Set To Begin Quantum Computing Evaluation This Year - QED-C. [online] Available at: https://quantumconsortium.org/almost-one-third-of-life-science-companies-set-to-begin-quantum-computing-evaluation-this-year
Qureca, 2020. Overview On Quantum Initiatives Worldwide - Qureca. [online] Available at: https://www.qureca.com/overview-on-quantum-initiatives-worldwide/
QuintessenceLabs, 2019. Breaking RSA Encryption – An Update On The State-Of-The-Art. [online] Available at: https://www.quintessencelabs.com/blog/breaking-rsa-encryption-update-state-art
RAND Corporation, 2020. Securing Communications in the Quantum Computing Age: Managing the Risks to Encryption. [online] Available at: https://www.rand.org/pubs/research_reports/RR3102.html. Also available in print form.
Science, 2020. Hartree-Fock on a superconducting qubit quantum computer. 369(6507), pp.1084-1089.
Simonite, T., 2018 What Is Quantum Computing? The Complete WIRED Guide. [online] Wired. Available at: https://www.wired.com/story/wired-guide-to-quantum-computing/
Singularity Hub, 2017. 6 Things Quantum Computers Will Be Incredibly Useful For. [online] Available at: https://singularityhub.com/2017/06/25/6-things-quantum-computers-will-be-incredibly-useful-for/#sm.00001itbsncl74elnykbatbf5ms2n
Standards.ieee.org, 2017. IEEE Approves Standards Project For Quantum Computing Definitions. [online] Available at: https://standards.ieee.org/news/2017/ieee_p7130.html
Standards.ieee.org. 2018. P7131 - Standard For Quantum Computing Performance Metrics & Performance Benchmarking. [online] Available at: https://standards.ieee.org/project/7131.html
The Quantum Daily, 2020. Apexqubit: Harnessing Quantum Computing For Health And Humanity. [online] Available at: https://thequantumdaily.com/2020/02/28/apexqubit-harnessing-quantum-computing-for-health-and-humanity
The Quantum Daily, 2020. TQD Exclusive: A Detailed Review Of Qubit Implementations For Quantum Computing. [online] Available at: https://thequantumdaily.com/2020/05/21/tqd-exclusive-a-detailed-review-of-qubit-implementations-for-quantum-computing/
Volkswagenag.com, 2018. Volkswagen Ensures Intelligent Traffic Management With Quantum Computers. [online] Available at: https://www.volkswagenag.com/en/news/2018/11/Volkswagen_intelligent_traffic_management.html#
Wallden, P. and Kashefi, E., 2019. Cyber security in the quantum era. Communications of the ACM, 62(4), pp.120-120.
Willis Towers Watson, 2019. Willis Towers Watson And Microsoft Announce Quantum Computing Collaboration To Transform Risk Management. [online] Available at: https://www.willistowerswatson.com/en-GB/News/2019/05/wtw-and-microsoft-announce-quantum-computing-collaboration-to-transform-risk-
management
World Economic Forum, 2020. Global Future Council On Quantum Computing: Frequently Asked Questions. [online] Available at: https://www.weforum.org/reports/global-future-council-on-quantum-computing-frequently-asked-questions
Legal disclaimer
This document is not intended for distribution to, or use by, any person or entity in any jurisdiction or country where such distribution or use would be contrary to local
law or regulation. It is the responsibility of any person communicating the contents of this document, or any part thereof, to ensure compliance with all applicable legal
and regulatory requirements.
The above documents do not represent a prospectus or invitation in connection with any solicitation of capital by Lloyd’s. Nor does it constitute an offer by Lloyd’s to
sell securities or insurance, a solicitation or an offer to buy securities or insurance, or a distribution of securities in the United States or to a U.S. person, or in any other
jurisdiction where it is contrary to local law. Such persons should inform themselves about and observe any applicable legal requirement.
The above documents have been produced by Lloyd's for general information purposes only. While care has been taken in gathering the data and preparing this
document, Lloyd's does not make any representations or warranties as to its accuracy or completeness and expressly excludes to the maximum extent permitted by
law all those that might otherwise be implied.
Lloyd's accepts no responsibility or liability for any loss or damage of any nature occasioned to any person as a result of acting or refraining from acting as a result of,
or in reliance on, any statement, fact, figure or expression of opinion or belief contained in this document. The content in these documents does not constitute advice of
any kind.
Classification: Confidential

The Impacts of Quantum Computing On Insurance: February 2021

Uploaded by

Copyright:

Available Formats

The Impacts of Quantum Computing On Insurance: February 2021

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Impacts of Quantum Computing On Insurance: February 2021

Uploaded by

Copyright:

Available Formats

The impacts of

The (not so) basics of quantum computing

Quantum computing’s threat to cyber security

The quantum landscape

Insurance impacts – navigating the quantum realm

Quantum computers within reach

The power of qubits

How much faster are quantum

Qubits are extremely sensitive to noise. The slightest

Decoherence means qubits are highly unstable and

What types of problem can a

Quantum computing has the potential to revolutionise

Quantum computing (QC) could be used to streamline

Fertiliser production Hydrogen production

Volkswagen, in partnership with D-Wave, are

Financial services applications

Dynamic portfolio Risk profiling and

The current state of quantum computing development is akin to

Error correction remains one of the most resource intensive and

The quantum computing ecosystem

For practical applications of quantum computing, we need

Encryption is used to securely transfer information between Demystifying encryption - How do we

The security of most online transactions and data held by

The difference with quantum computing adversaries is that

Quantum computing doesn’t equally affect all cryptosystems

Post-quantum cryptography (PQC) refers to classical algorithms

Case study: The quantum threat to

The digital signature that authenticates blockchain

In assessing whether we should be worried about the threat of

1. How soon are quantum computers capable of breaking

Source: (1), (2) RAND Corporation, 2020

Source: (1) Lloyd’s, 2019 (2) Lindsay, 2020

Business opportunities for insurers

Even though quantum computers capable of breaking

Operational benefits to insurers

Operational benefits to insurers:

Source: (1) Munich Re, 2019 (2) Nature, 2020

Moving forward – six actions you could take

Source: (1) Novarica, 2019

Source: (1) – (3) Novarica, 2019

Innovatorsunder35, 2018. Noor Shaker. [online] Available at: https://www.innovatorsunder35.com/the-list/noor-shaker/?platform=hootsuite

Science, 2020. Hartree-Fock on a superconducting qubit quantum computer. 369(6507), pp.1084-1089.

You might also like