Last year, we compiled our first annual report on AI governance. The purpose was to identify critical progress from numerous AI governance studies. We were pleasantly surprised by the enthusiastic responses to our invitations, resulting in 50 expert contributions to our report. The positive feedback from various individuals and organizations on the final publication encouraged us to continue the initiative in the future. Notable contributions include the recommendation of the Montreal AI Ethics Institute and a letter from the senior advisor at the Office of the President of the United Nations General Assembly. We hope that this report can improve understanding of - and help to bridge - different viewpoints on the challenges and opportunities of AI governance. That is the reason we compiled the report this year.
2020 will leave a deep mark in human history, as the outbreak and spread of the COVID-19 pandemic strained the economic and social development of the whole world. We once even expected that little progress on global AI governance would be made in 2020. However, there was still significant interest in continuing the annual report. Compared with last year, the number of participating authors (and institutions) turned out to be a little higher this year, as 52 experts (from 47 institutions) provided contributions.
As some authors have worked on this report for two consecutive years, they have been able to build on their work from the first year. Take OpenAI as an example: while its release plan for GPT-2 in 2019 sparked some controversies, the new release plan they proposed at the launch of GPT-3 in 2020 seems better received. The European Union is another good example: following its AI Ethical Framework released in 2019, it issued a White Paper in 2020, proposing corresponding regulatory rules.
The unusual situation created by the pandemic has also resulted in serious reflection on AI and its governance. Being compelled to reflect on AI may provide us with new ideas for future exploration. Seán ó héigeartaigh from the University of Cambridge has been delving into whether AI deserves its hype or whether attention should be focused on the basics of the problem, like investments in public health during the pandemic. Other experts question whether the numerous existing AI governance studies can be effectively translated into policies for dealing with the pandemic.
By deciding to put together this global observation report this year, we were also able to invite experts and institutions who were not involved in the previous year. After the previous report was released, it was pointed out that the voice of the Global South, especially Latin America and Africa, had been neglected. As a result, an effort was made to include experts from Latin America and Africa, to reflect the concerns and the work done in AI governance there.
To download the full report please click here: AI Governance in 2020