AI is no longer just moderating individual posts — it is learning how to interpret and enforce policy itself. Dave Wi llner — who has led trust and safety teams at Facebook\, Airbnb\, and OpenA I — joins journalist Meg Marco for a conversation about the shifting terrai n of moderation in the age of generative AI.
From the tipping points of past technologies\, like the cell phone\, to today’s new “policy-rewriti ng machines” that can run millions of classifications in an hour\, Willner and Marco will explore how automation is changing the scale\, speed\, and s takes of online governance.
New systems can ingest a company’s policy documents\, apply them to vast datasets\, and make millions of classificat ions in hours. In doing so\, they both replicate the human task of drawing boundaries — deciding what falls inside or outside the rules — and extend t hat task far beyond traditional concerns like hate speech or misinformation .
Dave Willner is a Co-Founder at Zentropi\, the company behind CoPE\, a best-in-class small language mode l capable of accurate and steerable content classification.
He was pr eviously a non-resident fellow in Stanford Cyber Policy Center\, and worked in industry as Head of Trust & Safety at OpenAI\, as Head of Community Pol icy at Airbnb\, and as Head of Content Policy at Meta (formerly Facebook).< /p>
Meg is the Senior Director of the Applied Social Me dia Lab\, focusing on building public interest technology that helps make i nformation available and understandable to researchers\, journalists\, civi l society organizations and the general public. She has held senior editori al positions at WIRED\, ProPublica and The Wall Street Journal.
CATEGORIES:Speaker/Panel LOCATION:Lewis Hall\, 5th floor at the Berkman Klein Center's Multi-Purpose Room 515 GEO:0.000000;0.000000 ORGANIZER;CN="Jessica Weaver":MAILTO:jweaver@law.harvard.edu URL;VALUE=URI:https://hls.harvard.edu/events/automating-content-policy/ ATTACH;FMTTYPE=image/png:https://hls.harvard.edu/wp-content/uploads/2025/10/1Speaker_16x9.png END:VEVENT BEGIN:VTIMEZONE TZID:America/New_York BEGIN:DAYLIGHT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 DTSTART:20250309T070000 TZNAME:EDT END:DAYLIGHT END:VTIMEZONE END:VCALENDAR