Marking a year since the introduction of ChatGPT, the two-day summit featured five members of Congress and dozens of leaders in research, industry, policy and law.
Authored by: Ruth Steinhardt | Read the original GW Today article.
On Nov. 30, 2022, OpenAI introduced its game-changing large language model ChatGPT to the public. A year later, global leaders in research, industry, thought and policy including multiple members of Congress convened at the George Washington University for the fifth edition of the Athens Roundtable on Artificial Intelligence and the Rule of Law, a summit on ethical AI development and governance.
Co-founded and sponsored by the nonprofit The Future Society,this year’s edition of the roundtable featured more than a dozen co-sponsors, including GW’S Institute for International Science and Technology Policy; NIST-NSF Institute for Trustworthy AI in Law & Society; the Embassy of Greece in Washington, D.C.; OECD; World Bank; Center for AI and Digital Policy; UNESCO; Homo Digitalis; IEEE; Paul, Weiss LLP; Arnold & Porter; and the Patrick J .Mcgovern Foundation. The event is an opportunity to share knowledge across disciplines and, through that dialogue, develop future-proof policies with real-world impact in a rapidly evolving field.
That mission aligns precisely with GW’s strengths and its institutional tradition of evidence-based policy impact, President Ellen M. Granberg said in introductory remarks Thursday at the Jack Morton Auditorium.
“We’re not an institution that is content with just publishing scholarship and hoping someone else will decide what to do with it,” Granberg said. “What makes GW unique is the way in which we extend our scholarship to direct applications across education, policy, patient care and other areas. The university’s location in the nation’s capital, combined with its diverse and highly talented faculty, can connect science, technology and innovation with law, policy and ethics like very few other institutions can across the globe. Together our students and faculty are working to find real solutions to some of society’s most pressing challenges.”
Featured speakers at the two-day event included U.S. Sens. Richard Blumenthal (D-Conn.), Amy Klobuchar (D-Minn.), Brian Schatz (D-Hi.) and U.S. Reps. Yvette Clarke (D-N.Y.) and Sara Jacobs (D-Calif.); representatives from the governments of Tanzania, the Czech Republic and others and from intergovernmental organizations including the European Union and the United Nations; industry leaders from Google and elsewhere; and researchers and academics from across the United States and the world.
U.S. lawmakers stressed the importance of bipartisan cooperation to create meaningful federal regulations for AI development and deployment, enabling innovation but preventing AI’s potentially catastrophic societal outcomes. That means such regulation needs to be nimble rather than purely reactive. Some areas of concern are already identifiable—data security, fraudulent AI-generated data, the electoral impact of “deepfakes”—while others will arise as these technologies develop.
“What we need are some basic, common sense, future-proof principles that set clear rules of the road to help developers and companies innovate responsibly while also protecting consumers from potential harms,” said Schatz, who has introduced legislation to label AI-generated content and to empower a federal commission to develop a regulatory structure for AI, much as the Communications Act did for radio and television in the 1930s and the Communications Decency Act did for the internet in the 1990s.
Klobuchar said the issue is of bipartisan concern, particularly when it comes to misinformation and fraud. She has partnered across the aisle with Sens. Susan Collins (R-Maine), Chris Coons (D-Del.) and Josh Hawley (R-Mo) to ban the use of AI to generate deceptive content influencing federal elections.
“Leaders from both sides of the aisle agree: We can’t sit on the sidelines while AI continues to advance,” Klobuchar said. “I really believe this is our moment to ensure that future generations around the world can take advantage of the benefits of AI without sacrificing their personal security or endangering our democracy.”
Legislative approaches to AI should also be based on a thorough understanding of the regulatory failures in the 2010s that led to a few monolithic corporations’ domination of the current social media landscape, the lawmakers said.
“Congress had a choice: Should we protect consumer privacy? Should we stop companies from amassing power?” Blumenthal said. “We all know how that story ended. Congress failed. It failed to act and now gigantic monopolies have disproportionate and info-rich power over huge segments of our economy and our law.”
GW has established itself as a leader in the AI space, particularly on questions of policy and ethical governance. The university co-leads the NIST-NSF $20 million Institute for Trustworthy AI in Law & Society (TRAiLS), which works to develop new AI technologies that mitigate risk and promote trust by empowering and educating the public.
GW faculty experts, including TRAiLS principal investigators Susan Ariel Aaronson and David Broniatowski and Institute for Data, Democracy and Policy Director Rebekah Tromble, participated in panels and conversations throughout the summit, as did Elliott School of International Affairs Dean Alyssa Ayres. Vice Provost for Research Pamela M. Norris delivered welcoming remarks on the second day of the event.
“We all understand that AI systems have great potential to increase productivity and to spur innovation. AI will touch every aspect of our lives,” Norris said. “But in our haste to realize these gains, conversations like this are critical to consider the questions of governance and the guardrails that may be necessary. We owe this to the next generation. GW is not only convening these conversations but shaping them.”