23 September 2026
Introduction: AI Alignment, Reward Hacking and Model Capabilities
Facilitator: Dr. Fırat Akova
AI Safety and Governance Program
A free 4-week online reading and discussion program introducing AI safety and governance. Wednesday evenings 19:30–21:00, 23 September – 14 October 2026.
The AI Safety and Governance Program will take place on Wednesday evenings, 19:30–21:00, on 23 September, 30 September, 7 October and 14 October. The program is introductory in nature, open to everyone and free of charge.
The first session covers the field’s core concepts. What does AI alignment mean? Why can models be fragile against small-scale attacks that poison their training data? How does reward hacking show up in today’s models?
In the second session, participants will meet the most prominent risks in safety and governance: power concentration, gradual disempowerment, engineered pandemics and the collapse of critical infrastructure.
The third week examines AI’s dizzying progress and what that progress means for the majority of the world.
The final session focuses on careers in AI safety and governance, making concrete how the conceptual ground built over the four weeks can connect to participants’ own career plans.
Dr. Fırat Akova, visiting researcher at GovAI, and Berke Çelik, co-founder of AI Safety Türkiye, will be speaking at the event. Working through readings from BlueDot, Anthropic and METR, the program covers the technical and governance sides together, and sessions are driven by participant discussion.
🔗 Register on Luma. Your spot is confirmed once your application is approved.
Program schedule
30 September 2026
Power Concentration, Gradual Disempowerment, Engineered Pandemics and Critical Infrastructure Collapse
Facilitator: Dr. Fırat Akova
Participants split into groups; everyone picks one of the readings below.
7 October 2026
What Happens to the Majority of the World? The Trajectory of Frontier Models
Facilitator: Berke Çelik
14 October 2026
AI Careers
Facilitator: Dr. Fırat Akova
Facilitators
Dr. Fırat Akova
Completed his PhD in Philosophy at the University of Warwick with a thesis on effective altruism. He is a researcher at the Centre for the Governance of AI (GovAI); previously he lectured in philosophy at Koç University, was an Oxford Global Priorities Fellow, and researched AI sentience and digital minds at Future Impact Group.
Berke Çelik
Berke Çelik is the co-founder of AI Safety Türkiye, which runs education programs and community work on AI safety in Türkiye. Earlier work includes contributing to AI governance research and running the Development Program at the Global Policy Research Group, as well as roles at the Lead Exposure Elimination Project and the Alliance for Reducing Microbial Resistance. Berke studied Philosophy at Boğaziçi University.
AI Safety and Governance Program
Open on Luma