About workshop
The presentations will be followed by a debate at the end, where we invite the audience to present questions on the YouTube channel for the transmission (youtube.com/nurep_ext).
We are now faced with an unprecedented reality, as the very act of creation, which lies at the very core of intellectual property as the protection of human genius, is no longer exclusive to the human being. When it comes to artistic and literary works, we already observe AI technologies that allow for the development of products that, if created by human beings, would be protected by copyright. To further complicate the issue, there are numerous combinations between different degrees of human interference, system autonomy and predictability of the final results – which calls into question the adequacy of the current copyright regimes for AI-generated works and their overarching implications for a legal system that regards human beings as their sole point of departure.
As discussions on the applications of artificial intelligence systems grow in intensity and frequency, special care should be devoted to the handling of inputs and outputs of such technology. More specifically, on this workshop we direct our attention to the legal protection granted to compilations of data and to the materials resulting from the work performed by machine learning mechanisms – both of which lie, in great measure, in the realm of intellectual property in general and more specifically within copyright.
Concerning compilations of data, the absence of a single model in the world regarding the appropriation, regulation or the implementation of public policies dedicated to define the allocation of property rights and possible uses of such materials by third parties in a lawful manner creates a considerable risk of stifling innovation and deepening economic, technological and social rifts between countries at different stages of development. Such issue has only been exacerbated by the recent COVID-19 crisis and invites a renewed observation.
Access to databases and their content in order to extract data is not formally permitted in the overwhelmingly majority of national legal systems nor is it regulated at the international level. Therefore, most processes of extracting and reusing data from the different databases in order to train AI systems are not in accordance to the current copyright legislation. If legal action is taken by the database rights holders, the economic consequences may be severe for the infringer. More broadly it may have a chilling effect on innovation for data intensive industries.
Considering the growing relevance of the issue at hand and the possibilities for exchanging experiences and knowledge within the research community, we propose for ICAIL 2021 a workshop focused on the question of copyright regulation and how it affects artificial intelligence, especially in relation to two aspects: copyright treatment of data and text mining techniques and the discussion of current forms of protection granted to AI-generated artistic-like works.
The workshop consists of three presentations followed by a panel discussion at the end, where we invite the audience to ask questions and interact with presenters. The proposed length is of half of one day.
The first presentation shall be focused on the adequacy of current copyright architecture for works generated, entirely or partially, by artificial intelligence systems, and will be presented by Dr. Luca Schirru, whose research revolves around the workings of AI,their products and their implications for copyright.
The second presentation will turn our attention to the construction of copyright limitations and exceptions for text and data mining techniques (TDM) around the world. This presentation shall be performed by MSc. and PhD candidate Miguel Alvarenga, who researches database protection, data and text mining permissible uses and their effects on innovation and scientific research.
Finally, the third presentation will be delivered by Professor Allan Rocha de Souza and will concentrate on the pillars of copyright protection, the limitations and exceptions opposed to it and the contemporary challenges to its structure and fundamentals, both at the national and international levels, as well as elaborate on possible new models for protection that takes into account its potential negative effects on research and innovation.
At the end of the presentations, we will hold a discussion panel with the three presenters under the coordination of Dra. Patricia Porto, where the audience will have the broad possibility to intervene, question and provoke the speakers on the issues at hand.
This workshop is proposed and organized by NUREP – Núcleo de Pesquisa em Direitos Fundamentais, Relações Privadas e Políticas Públicas (Research Group on Fundamental Rights, Private Relations and Public Policy – http://dgp.cnpq.br/dgp/espelhogrupo/3462805150382122), which is lead by Professor Allan Rocha de Souza and is directly associated to the Federal University of Rio de Janeiro Graduate Program on Public Policy, Strategies and Development (UFRJ/PPED) and the Federal Rural University of Rio de Janeiro Law School (UFRRJ/ITR).
It is intended for a half day and the audience is expected to be made of law and public policy professionals and students, researchers, creative industries business people, data scientists, government representatives and non-governmental organizations. It will be publicized in academic and professional networks, government agencies and among NGOs. Key players will be directly invited to take part of the workshop.
Format
The workshop includes an open call for papers, public online presentations by 03 guest speakers on specific issues, and the presentation of up to 06 selected papers, whereby each presentation (guests and selected papers) will be of 20 minutes, followed by an one-hour open discussion at the end, where we invite all the presenters and the audience to ask questions and interact.
Speakers and Moderators
- Allan Rocha de Souza
- Patrícia Porto
- Luca Schirru
- Miguel Alvarenga
Dates
All papers shall be sent exclusively to the following e-mail aicopyrighticail21@