TRAIF Preview 2020: Looking through a kaleidoscope of AI ethics

The Responsible AI Forum Preview took place from November 12th to 13th, 2020. Over 20 global experts affiliated with the IEAI and its networks joined virtually to discuss their cutting-edge research and findings around some of the most relevant issues in the field of AI Ethics. The event gave the audience a glimpse into some of the most pressing issues concerning the field of AI Ethics and the efforts of the IEAI, together with its network members and partners, to address such important issues and achieve applicable solutions that can be embedded in the design, development and deployment of AI systems.

Session 1

The participants of this session were the winners of the 2020 Call for Proposals for multi-disciplinary research on ethics of the use of AI to manage pandemics and health-related crises. In this session the moderator was the director of the IEAI, Christoph Lütge, and the participants were Dirk Brand, Christian Djeffal, Mark Findlay, Jeannie Paterson and Stefaan Verhulst. The panelists discussed their research projects and approaches for developing clear instructions on embedding AI ethics principles in practical ways into AI-enabled systems used to manage health crises, such as the current COVID-19 pandemic. All of these cutting-edge research initiatives take a multi-disciplinary and multi-level approach enabling an effective assessment of relevant technologies. The goal of these projects is to provide recommendations that can be implemented in AI tools used in managing health crises in order to capitalize upon the potential of AI in helping to face global challenges, while at the same time safeguarding the well-being of humans.

Session 2

This session was devoted to exploring some of the IEAI’s research clusters concerning AI Ethics in the context of online behavior and choice autonomy. The moderators were Matthias Uhl and Caitlin Corrigan, and the distinguished guests of this session were Anna Baumert, Jens Grossklags, Irina Dolgopolova and Jürgen Pfeffer. The panel members stem from different research areas but were working together in interdisciplinary teams to explore important issues concerning the mitigation of potentially harmful online trends.

The talks and discussion explored the need of effective tools that can prevent, predict and moderate online behavior, as well as the transparency needed for the users’ perception of moderating systems.

The preliminary results of the research projects have already displayed the importance of addressing these issues, the need to analyze how AI enabled algorithms can create and reinforce new and potentially harmful dynamics and what practical solutions can be developed to detect such dynamics and mitigate their consequences.

Session 3

This session was devoted to sharing the work of the Responsible AI Network – Africa (RAIN-Africa), an initiative founded by the TUM IEAI and Kwame Nkrumah University of Science and Technology (KNUST) in Ghana. The moderators were the scientific coordinator of the IEAI, Caitlin Corrigan, and Jerry Kponyo, co-founder of RAIN-Africa. The distinguished guests of this session were Paa-Kwesi Gadagbui, Teki Akuetteh Falconer, Abraham Kuuku Sam and Kwame Osei Boateng. The panel members discussed the potential benefits and risks of the development and deployment of AI systems in Africa, particularly focusing on developing a strategy for long-term sustainability. The researchers shared their investigations into the challenges concerning the field of AI ethics on the African continent by exploring how accountability, regulations, standards and frameworks, need to be implemented. They considered challenges such as the technical know-how on AI, the access to data and technologies, as well as the policy frameworks for ethical AI in the African contexts. They raised a number of important questions that need to be addressed in order to capitalize upon the opportunities for AI-enabled tools to escalate initiatives for sustainable development on the African continent.

Session 4

During this session, several members of the Global AI Ethics Consortium (GAIEC) discussed the role of exchange and partnerships in promoting the harmonization and sharing of global perspectives on AI Ethics. The IEAI has continued to grow its network by connecting currently 20 experts from 7 continents who are together addressing some of the biggest global challenges in AI ethics. This session was moderated by Caitlin Corrigan and the distinguished members of the panel included Christoph Lütge, Rafael Calvo, Jean-Gabriel Ganascia and Yi Zeng. The panel members explored the importance of independent research conducted in academia and the importance of working together with international bodies, such as NGOs and governments, in order to contribute to the refinement of frameworks at a global level. This session was particularly interesting due to the international nature of the panel and their discussions, which highlighted emerging challenges with AI systems deployed in different societal contexts.

Session 5

The final session of the event gave the audience a glimpse into the research clusters of the IEAI concerning ethics of AI-enabled decision-making practices. This session was moderated by Christoph Lütge, and the panel speakers of this session were Maximilian Geißlinger, Franziska Poszler, Charlotte Haid, Mrinalini Kochupillai and Lukas Meier. The speakers are researchers affiliated with the IEAI exploring the use of AI in areas such as mobility and safety, future of work, sustainability as well as medicine and health care. This session particularly highlighted how difficult it is to implement and embed AI ethics guidelines into systems and, at the same time, how necessary it is to address the everyday ethical challenges confronted when using AI-enabled tools in order to obtain applicable solutions to these problems. The panel members presented some of the preliminary results of their research projects, which are all centered on achieving applicable solutions to dilemmas emerging in the use of AI in decision- making contexts.

The TRAIF Preview was an intriguing teaser for The Responsible AI Forum 2021 (TRAIF 2021), which is hosted by the IEAI and will take place in 2021 in Munich. This international Forum aims to bring together members from academia, civil society, government and industry in order to discuss the most relevant issues concerning the field of AI ethics and explore how the cutting-edge research conducted in this field can obtain practical applications that can be embedded into AI systems. Check out the website of the TRAIF 2021 on www.responsibleaiforum.com  and register for the three-day event by submitting proposals for papers, workshop and roundtables. Our call for proposals is still open until the 31st January 2021.