Press & Talks
Public work across journalism, policy rooms, research labs, and technical conferences.
Selected from 60+ invited talks, lectures, keynotes, panels, and podcast appearances.
Press
Selected coverage and quoted pieces.
More People Are Using AI Chatbots for Health and Relationship Advice, Study Shows
AI Observatory
2026The Washington PostNew Data Shows People Are Getting Personal With AI
AI Observatory
2026MIT Technology ReviewWe Still Don't Know How People Are Really Using AI
AI Observatory
2026WiredYou Can Now Sound the Alarm on AI Behaving Badly
FLARE-AI
2026The Wall Street JournalWorkers Sue $10 Billion AI Startup for Collecting and Exposing Personal Data
AI data and labor
2026MIT Technology ReviewWhat's Next for Chinese Open-source AI
Open model ecosystems
2026Financial TimesChina Leapfrogs US in Global Market for 'Open' AI Models
Open model ecosystems
2025TechCrunchStudy Accuses LM Arena of Helping Top AI Labs Game Its Benchmark
Leaderboard Illusion
2025Ars TechnicaResearchers Claim LM Arena's AI Leaderboard Is Biased Against Open Models
Leaderboard Illusion
2025WiredResearchers Propose a Better Way to Report Dangerous AI Flaws
AI flaw disclosure
2025The New York TimesThe Data That Powers A.I. Is Disappearing Fast
Consent in Crisis
2024The Washington PostTop AI Researchers Say OpenAI, Meta Hinder Independent Evaluations
Safe Harbor
2024MIT Technology ReviewThis Is Where the Data to Build AI Comes From
Multimodal Data Provenance
2024The Washington PostAI Researchers Uncover Ethical, Legal Risks to Using Popular Data Sets
Data Provenance Initiative
2023The New York TimesMaybe We Will Finally Learn More About How AI Works
Foundation Model Transparency Index
2023The AtlanticWe Don't Actually Know If AI Is Taking Over Everything
Foundation Model Transparency Index
2023Talks
Selected invited talks, keynotes, panels, and podcasts.
Y Combinator Paper Club
ATLAS
Invited Speaker
UC Berkeley CS 294/288
ATLAS
Guest Lecturer · Hosted by Sewon Min
Stanford UN AI for Good Law Track
Benchmarking & Auditing AI Systems in High-Risk Domains
Panelist
Cohere Labs Connect
Research Collaborations
Keynote Speaker
Knight First Amendment Institute
AI and Democratic Freedoms
Panel Moderator
ICLR 2025 Building Trust in LLMs Workshop
AI Coordinated Disclosure
Invited Speaker · Third-party AI evaluation
ICLR 2025 Open Science for Foundation Models
Invited Speaker · Open data and data provenance
BBC Radio 4
Will AI Eat Itself?
Podcast Guest · Consent in Crisis
MIT Imagination in Action
Invited Speaker · Data Provenance Initiative
USC NLG Seminar Series
Invited Speaker · Data Provenance Initiative
Microsoft Research India
Invited Speaker · A Pretrainer's Guide