Michal Kosinski

Michal Kosinski is one of the world's most influential academics working on artificial intelligence. A Professor of Management and Strategy at the University of Hong Kong - previously at Stanford University and the University of Cambridge - he uses AI to study people, and psychology to study AI.

Michal has keynoted for organizations including PwC, Google, McKinsey, Boston Consulting Group, P&G, and Deutsche Telekom, and has spoken at venues ranging from the European Parliament and Chatham House to C-SPAN and DLD - where his 2025 talk on the psychology of AI was voted Best Talk of the Day. He advises governments and corporations on AI, having testified before the FTC, the Department of Justice, and the EEOC, and serves on the AI Board of TELUS. Audiences rate him exceptionally: 9.81/10 at a recent leadership event - "the highest score I've ever seen," per the organizer.

He was among the first to show human-like psychological processes emerging in large language models — theory of mind, cognitive reflection, reasoning no one designed into them — research that changed how scientists and companies think about what AI can do. His earlier work established the reverse direction: that ordinary digital footprints — likes, clicks, images — reveal our intimate traits. He was behind the first press article warning against Cambridge Analytica, published three years before the scandal broke.

One of the world's most cited scientists (30,000+ citations, Clarivate top 1%), his research has inspired a cover story in The Economist, the theatre production "Privacy" (Donmar Warehouse, London; PublicTheater, New York), several TED talks, and a video game - and has been featured in thousands of press articles, podcasts, and documentaries worldwide.

Keynote topics:

• Making Sense of Modern AI (core session)
  How today's AI systems actually work, what they are genuinely good at,and where they reliably fail. Executives leave with a clear, practicalmental model of AI - and sharper judgment about when to trust it, howto validate it, and where human leadership remains critical.
• Digital Twins: Predicting and Modeling Human Behavior with AI
  Digital records enable algorithmic models of individuals' traits,preferences, and likely behaviors, with powerful applications inleadership, talent, marketing, and healthcare - and serious risks ofmanipulation and surveillance. Leaders leave knowing what these modelscan reliably predict, where their limits lie, and where the red linesare.
• The End of Privacy? Psychological Profiling in the Digital Age
  Digital traces predict personality, values, abilities, and preferences- often better than traditional assessments, and often withoutpeople's awareness. Leaders leave understanding what profiling can andcannot do, and how to harness it while avoiding ethical, legal, andreputational disasters.

Sessions work as keynotes, executive-education sessions, or boardbriefings - standalone or combined into a short program.