March 28, 2025
Columbia University and NYC Co-Sponsors* hosted another PCRI Symposium on March 28, 2025. This day-long virtual symposium addressed various issues pertaining to research integrity in the age of generative AI.
March 28, 2025 Videos
Introductory Remarks given by Jeannette Wing (Executive Vice President for Research)
Session 1: Fireside Chat | The Revolutionary Power of Gen. AI in the Scientific Domain
This session featured Marcia McNutt, President of the National Academy of Sciences, and Eric Horvitz, Chief. Scientist, Microsoft, a pioneer in artificial intelligence, as they explore the revolutionary role of Generative AI in science. This session highlighted how AI is transforming research practices and policies, driving innovation, and expanded the boundaries of what is scientifically possible, but also presented new risks, challenges and ethical quandaries in relation to the integrity of science.
This interactive panel discussion explored the "inside story" of a hypothetical research misconduct case involving questionable data in a high-profile oncology publication. Through a five-scene presentation, panelists portrayed key stakeholders— that included a principal investigator, graduate student, research integrity officer, journalist as they navigated the complex challenges that emerged when scientific misconduct is suspected and investigated. The session examined critical decision points from multiple perspectives.
Moderator: Naomi Schrag
Panelists: Ai Yamamoto, Maria Blumenkrantz, Reginald Miller, Susan Garfinkel, and Mark Hansen
In this session, journal editors explored the complex challenges editors faced when confronting problematic publications—from coordinated paper mills to image manipulation and AI-generated content. Panelists discussed practical strategies they employed to detect and address research misconduct, including their approaches to digital image verification, artificial intelligence detection, and authorship disputes. They also discussed peer review, shared insights on managing relationships with external “data sleuths,” and described how journals of different sizes navigated current challenges.
Moderator: Arturo Casadevall
Panelists: Kaia Motter, Sri Devi Narasimhan, and Lawrence Lustig
This session brought together researchers from various disciplines to discuss the latest innovations in Generative AI and their impact on their respective fields. The session consisted of 10-minute mini-talks that discussed a particular GAI use case. Following the presentations, a moderated discussion addressed the steps the researchers were taking to ensure the credibility of the models they were building, how they addressed potential bias, and the broader ramifications of using GAI in their fields and research integrity.
Moderator: Frank Pasquale
Panelists: Siwei Lyu, Kavita Bala, and Despina Kontos
Speakers Included:
Eric Horvitz, Chief Scientific Officer, Microsoft
Marcia McNutt, President, National Academy of Sciences
Jeannette M. Wing, Exec. Vice Pres. for Research, Columbia University
Kaia Motter, Head of Academic Affairs, Springer Nature
Arturo Casadevall, Johns Hopkins University, Bloomberg School of Public Health
Lawrence Lustig, Columbia University Irving Medical Center
Frank Pasquale, Cornell Tech
Siwei Lyu, SUNY Buffalo
Sri Devi Narasimhan, Cell Press
Mark Hansen, Columbia University
Despina Kontos, Columbia University Irving Medical Center
Ai Yamamoto, Columbia University Irving Medical Center
Kavita Bala, Cornell University
Maria Blumenkrantz, Columbia University Irving Medical Center
PCRI Symposium
Resources
NYC Cosponsors: New York University, Cornell University, including Weill Cornell Medicine and Cornell Tech, Memorial Sloan Kettering Cancer Center, The Rockefeller University, City University of New York, Icahn School of Medicine at Mount Sinai, Columbia University School of Nursing, Office of AI and Center for Community-Engaged Health Informatics and Data Science, Columbia University Fu Foundation School of Engineering and Applied Science, Columbia University School of Arts and Sciences, and Columbia University Data Science Institute
Questions? Contact the organizers at [email protected] | Follow the conversation #PCRISymposium