Experts Tackle Generative AI Ethics and Governance at 2025 K&L Gates–CMU Conference

DJ Patil, general partner of GreatPoint Ventures and former U.S. chief data scientist delivered a keynote address and held a fireside chat with Theresa Mayer(opens in new window), CMU vice president for research, on the conference’s first day.

Atoosa Kasirzadeh(opens in new window), assistant professor of philosophy in the Dietrich College of Humanities and Social Sciences(opens in new window), spoke on a panel that explored the new ethical concerns that arise with GenAI compared to conventional predictive AI. “The question of what are the measurement challenges when we want to enforce and implement AI risk governance frameworks are quite important,” she said “I hope that in the next year you are going to see much more mature analysis of this question.” She was joined on stage by David Danks of the University of California, San Diego; Francesca Rossi, IBM AI ethics global leader; and panel moderator Sina Fazelpour of Northeastern University. 

Mike Doyle, former U.S. representative from Pennsylvania’s 18th congressional district and government affairs counselor at K&L Gates, moderated a panel on existing government policies and guidance on GenAI with Jesse Dunietz, AI standards, policy and international engagement lead for the National Institute of Standards and Technology; Marc Rotenberg, executive director of the Center for AI and Digital Policy; and Max Katz, policy adviser for U.S. Sen. Martin Heinrich. 

Lorrie Cranor(opens in new window), FORE Systems University Professor of Computer Science and Engineering and Public Policy and director and Bosch Distinguished Professor in Security and Privacy Technologies at CMU’S CyLab Security and Privacy Institute(opens in new window) moderated a debate exploring internal organizational government structures in major AI firms.

“If we’re going to have systems that can replace what a person could do over a year, that opens up a mind-boggling array of possibilities in terms of risk,” observed Zico Kolter(opens in new window), professor and director of the machine learning department(opens in new window). “These questions still feel a  little like science fiction. But these are genuine concerns and we don’t have as much time as we think before they come to pass.” 

Provost James H Garrett Jr. (opens in new window)opened day two of the conference, which focused on the practical implementation of AI governance and the concrete impacts of the technology that are already manifesting themselves in education, health care, the workforce and beyond. 

“The theme of AI ethics and governance deeply aligns with Carnegie Mellon’s institutional mission,” he said. “We’re committed to not only advancing technological frontiers, but also ensuring these advancements serve humanity ethically and responsibly.” 

Carolyn Austin, director of practice innovation at K&L Gates, said while GenAI is being increasingly adopted by lawyers and law firms to enhance productivity, it does not absolve them of their legal and ethical responsibilities. “Lawyers have duties, including confidentiality, clear communication with clients, and partners have a duty to supervise lawyers and others reporting to them in their teams when they are carrying out the work, including appropriate use of generative AI technology. On the flip side, lawyers also have a duty of competence, which explicitly includes understanding the capacity and limitations of technology, including GenAI, and periodically updating that understanding. We can’t opt out. All of which underlines the need for thoughtful, internal AI governance policies and processes and for regular, thorough and targeted GenAI education.” 

Richard Scheines(opens in new window), dean of Dietrich College, moderated a fireside chat with Natasha Crampton, chief responsible AI officer for Microsoft. “First, it is the case, and this is very much my experience, that good governance is actually just good business,” Crampton said in response to a question about the potential for financial incentives to lead to unethical behavior in the AI industry. “But this is why, you know, it should not just be an industry endeavor, right? This is where academia plays a very important role. Governments play a very important role. Civil society plays an important role. There has to be a coming together of incentives to drive action across all of those players.”

Moderator Jodi Forlizzi(opens in new window), professor in computer science and CMU’s Human-Computer Interaction Institute(opens in new window) (HCII), led a conversation on the impacts of GenAI in education, employment, medicine and the environment with Sarah Fox(opens in new window), assistant professor in HCII; Marsha Lovett(opens in new window), vice provost for teaching and learning innovation at CMU; and Shiv Rao, founder and CEO of Abridge. 

Forlizzi asked Lovett about how to best incorporate AI when training new employees. She said it is important to examine both performance and learning outcomes.

“A learner may be able to perform better with the aid of an AI tool. It's a totally different question whether they have learned more with that AI tool,” Lovett said.