In February 2022 we published the second edition of our report Artificial Intelligence and Democratic Values. We expanded our coverage from 30 countries to 50. We also modified our metrics to take account of the new UNESCO Recommendation on the Ethics of AI. More than 100 members of the CAIDP Research Group made contributions to report.
We published the report in both PDF and EPUB formats, and discussed the findings at several international conferences devoted to AI policy.
Our AI Policy Clinic grew in 2022. More than 100 people applied for our introductory AI policy course and about 60 received formal certification. Many of our earlier graduates returned as Team Leaders and also to participate in our AI policy advanced course. Our original class took on new responsibilities as Regional Coordinators. We added a new AI policy seminar to explore the foundations of AI policy with detailed reading assignments. And we developed extern relations where students receive academic credit with both Stanford Law and Duke Law, in addition to Georgetown Law, our original partner. Our April 2022 Graduation featured remarks by Thomas Schneider, the newly elected chair of the Committee on AI of the Council of Europe.
The CAIDP Global Academic Network now includes over 30 leading experts in AI policy from almost 20 countries. The members of the GAN help guide the work of CAIDP, providing advice and direction, and also reviewing CAIDP program activities.
CAIDP is pursuing a wide range of projects on Artificial Intelligence policy with key partners
Since publication of the Center for AI Digital Policy (CAIDP) report Artificial Intelligence and Democratic Values and the award to European Commission President Ursula von der Leyen in mid-December, we have been busy engaging with policy makers, launching new projects, and pursuing our mission.
Here is a quick summary of work to date
AI and Fundamental Rights -
with Merve Hickok, MEP Alexandra Geese, Peggy Valcke, Oreste Pollicino, and Friederike Reinhold (CPDP 2021).
An International Accord for AI -
with Malavika Jayaram, Marc Rotenberg, MEP Eva Kaili, Marit Hansen, and Tuan Nguyen (CPDP 2021).
One of our big projects this year is the launch of the first AI Policy Clinic. The AI Policy Clinic is a cornerstone for the development of best practices and collaboration with other thinks tanks and university partners.
Our goal is train the next generation of AI policy experts. Participants acquire the basic skills of AI policy research, analysis, and advocacy in a formal academic setting. The class includes a syllabus, formal assignments, and mechanisms for evaluation. The AI Policy Clinic builds on the earlier teaching experience of the CAIDP staff as well as the pedagogy of law school school clinics. Key texts include the AI Policy Sourcebook (authored by Marc Rotenberg, CAIDP Executive Director) and the CAIDP report Artificial Intelligence and Democratic Values.
We use the phrase "AI Policy" purposefully to focus on the actions of governments, the development of law, and the role of advocacy organizations. That is consistent with the mission of the Center for AI and Digital Policy. As a consequence, we are generally avoiding "AI and Ethics," "AI and Research," "AI and Society," and all the "AI and <fill in key term>." Our interest and expertise is in the law and policy domain.
The AI Policy class includes students from the LLM Technology and Law program of Georgetown Law (USA), and the University of Basil (Switzerland), affiliated with the Michael Dukakis Institute.
Current projects include: comments to the European Commission on a proposal to ban biometric identification for mass surveillance, comments to the US National Security Commission on AI, "friend of the court" comments to the Court of Justice of the European Union concerning a case about AI used for border control, and weekly review of significant AI policy events worldwide. Once the Spring 2021 AI Policy Clinic is completed, we plan to publish an academic article describing the class, what we attempted to accomplish, and what others may learn from the experience.
We continue to publish our weekly newsletter, the CAIDP Update, to provide timely information about AI Policy. Recent CAIDP Updates reported on the remarks of President Joe Biden regarding AI policy at the Munich Security Conference (CAIDP Update 2.07), the decision of the Canadian Privacy Commissioners regarding ClearviewAI (CAIDP Update 2.06), the Council of Europe's proposed ban on facial recognition (CAIDP Update 2.05), President von der Leyen's welcome to the Biden administration and a proposal for a common digital agenda (CAIDP Update 2.04), the decision of the FTC to require the deletion of AI models developed from data unfairly obtained (CAIDP Update 2.03), a decision by an Italian court concerning the use of algorithms for employee determinations (CAIDP 2.02), and AI policy issues to watch in 2021 (CAIDP 2.01).
We are planning to expand our coverage of AI policy issues with a monthly survey that will look more closely at AI policy developments in six regions of the world - South America, North America, Africa, Europe, Central Asia, and East Asia.
CAIDP "booth" at the 2021 conference on Computers, Privacy and Data Protection. The organizers did an amazing job organizing an international conference with more than 1,200 people and 200 speakers from 80 countries, on gather.town.
CAIDP participated in the conference on Computers Privacy and Data Protection at the end of January and the OECD AI Workshop in early February. At CPDP, the CAIDP organized two panels - one on AI and Fundamental Rights, and the other on an International Accord on AI.
CAIDP Senior Researcher Merve Hickok chaired the first CPDP panel on "Fundamental Rights and AI." Panelists included MEP Alexendra Geese, CAHAI’s Peggy Valcke, Friederike Reinhold from AlgorithmWatch, and Professor Oreste Pollicino, representing Italy with the Fundamental Rights Agency and the Global Partnership on AI, also a member of the CAIDP team.
CAIDP Executive Director Marc Rotenberg chaired the second CPDP panel on an "International Accord for AI," which featured Michael Dukakis Institute CEO Tuan Nguyen, Member of the European Parliament Eva Kaili, German Data Protection Authority DPA Marit Hansen, and Digital Asia Hub founder and CAIDP Team member Malavika Jayaram.
Several members of the CAIDP team, including Larissa Zutter and Giuliano Borter also "managed our booth" at the CPDP 2021 gather.town exhibit hall.
Videos of CPDP 2021 panels are here.
Marc Rotenberg also spoke at the opening panel session for the OECD conference AI in Work, Innovation, Productivity and Skills. Marc highlighted the CAIDP report Artificial Intelligence and Democratic Values, and urged OECD countries to take further steps to implement the OECD AI Principles.
On February 19, the Michael Dukakis Institute organized the first panel on the International Accord on AI with several members of the CAIDP Team, including Governor Dukakis, Douglas Franz, and Merve Hickok, as well as Professor Nazli Choucri, President Vaira Vike-Freiberga (Latvia), Prime Minister Zlatko (Republic of Bosnia and Herzegovina), State Minister Yasuhide Nakayama (Japan), and the OECD’s Andrew Wyckoff.
The panel of world leaders and distinguished thinkers identified challenges, opportunities, and imperatives for the International Accord on AI. At the Michael Dukakis Institute Peace and Security Awards event in December 2020, European Commission President Ursula von der Leyen called for a Transatlantic Accord on AI.
CAIDP has established a methodology to evaluate national AI policies and practices. We have also gathered data to provide formal assessments and to rate and rank countries.
Now we are providing specific advice to countries regarding AI strategy. In late February 2021, we advised the US National Commission on AI regarding the draft final report for the US Congress and the US President.
CAIDP also provided advice to the European Parliament as it undertook a series of hearing on the External Dimensions of AI policy, under the direction of the Special Committee on AI in the Digital Age (AIDA).
In these policy recommendations, the CAIDP draws on the landmark reports Artificial Intelligence and Democratic Values, the CAIDP Updates, the expertise of the CAIDP Team, and recent developments. CAIDP is available to provide expert assessments for AI policies for those countries which we have reviewed.
One of CAIDP's new initiatives aims to promote public participation in the AI policy process. We are specifically promoting efforts by government to engage the public in AI decision making. Our hope is that the "Public Voice" will help promote better informed, more legitimate AI policies.
This initiative follows from our earlier determination that opportunities for public participation in AI policymaking is one of the key indicators of human-centric, trustworthy AI policy. In our 2020 report Artificial Intelligence and Democratic Values we rated countries more highly if they created opportunities for meaningful public comment.
One of the current priorities for the CAIDP is to develop significant policy inputs for the G20 meeting, which will take place later this year in Rome . The G20 is one of the leading organizations in the world of AI policy. In 2019, the G20 nations endorsed the OECD AI Principles, the most comprehensive framework for AI policy. Then in 2020, the G20 Digital Economy Task Force published a very good report, Examples of National AI Policies, which surveyed steps taken by governments to implement the OECD/G20 AI Principles. We relied on that report as we prepared the CAIDP report Artificial Intelligence and Democratic Values. In 2020, the G20 leaders also published a statement on Digital Policy that we consider very favorable.
So, in 2021 the CAIDP will be working with the G20 to (1) assess implementation of the OECD/G20 AI Principles, (2) carry forward the work of the report on National AI Examples, (3) and promote an even better declaration at the Leaders Summit in October. We will also engage on the special AI policy topics for 2021, such as the use of AI for public services. We have begun the process of meeting with the DETF and participating in the Working Groups for Civil Society (the "C20") and Think Tanks ("T20"). We are also collaborating with academic experts and NGOs in Italy, which is the host for the G20 in 2021.
Since 1949, the Council of Europe has promoted respect for human rights, democracy and the rule of law in its 47 member states, 27 of which are also members of the European Union. In September 2019, the Committee of Ministers of the Council of Europe set up the Ad Hoc Committee on Artificial Intelligence – CAHAI
The CAHAI is tasked with examining, through broad consultations, the feasibility and elements of a legal framework for the development, design and application of artificial intelligence, based on the Council of Europe’s standards on human rights, democracy and the rule of law.
On February 26, 2021, the CAIDP was granted Observer Status with the CAHAI.
CAIDP has worked closely with the Global Privacy Assembly, an international network of privacy officials and advocates. CAIDP cited the Global Privacy Assembly's resolutions on AI and Ethics, and AI and Accountability in the report Artificial Intelligence and Democratic Values.
CAIDP is now seeking observer status with Global Privacy Assembly.
CAIDP is working with the organizers of the excellent conference Computers, Freedom, and Data Protection to establish a global track for CPDP in 2022 that will enable participation by people around the world. CAIDP proposes specifically to focus on the Asia-Pacific and to build upon CPDP's already very successful model for the EU and North American participants.
As we move forward the work of CAIDP on AI policy through the AI policy Clinic, our publications, and our formal statements for governments, we are also building new connections with others around the world engaged in similar work. Many universities and think tanks are beginning the process of formalizing their own understanding of AI policy issues. We believe that the Center for AI and Digital Policy can play an important role in this process.
The aim of the CAIPD Academic Network for AI Law and Policy is to share relevant expertise and information, foster capacity-building and develop common policy and research projects. CAIDP has already published two substantial reference works on AI policy and will continue to publish the CAIDP Updates as well as provide related information for our academic partners.
For more information and to become a member, please contact CAIDP Global Program Director Karine Caunes.
If you like the work we are pursuing, there are are several ways to become involved:
We are very grateful for our amazing team at the Center for AI and Digital Policy. In just a few months, we have brought together many exceptional people and produced a lot of amazing work.
Thank you for your interest in the work of CAIDP!
Marc, Tuan, Anne, Karine, and Merve
CAIDP Senior Management Team