Cooperative AI Summer School
About
The Cooperative AI Summer School is an annual intensive educational program organized by the Cooperative AI Foundation (CAIF), a registered charity in England and Wales (charity number 1201294). The summer school was first held in July 2023 in London, UK, and has run annually since, with subsequent editions in Santa Cruz, California (2024), Marlow near London (2025), and Canada (2026). The program is designed to provide students and early-career professionals in AI, computer science, and related disciplines with a firm grounding in the emerging field of cooperative AI. The curriculum spans from foundational concepts to cutting-edge research, covering topics such as multi-agent reinforcement learning, game theory, fairness in AI, mechanism design, and language model cooperation. Lectures are delivered by prominent researchers from leading universities and AI labs worldwide. Each edition runs for approximately five days and is held in-person only, though lecture recordings are subsequently made available on CAIF's YouTube channel. Fees are tiered at roughly £250-£275 for students and independent researchers and £500-£550 for faculty and working professionals, with the foundation committed to providing fee waivers and travel bursaries to ensure no qualified applicant is excluded due to financial constraints. The summer school is directed by Lewis Hammond (Research Director at CAIF) and Cecilia Elena Tilli (Associate Director of Research and Grants at CAIF). The Cooperative AI Foundation itself is backed by a $15 million philanthropic commitment from Macroscopic Ventures and has a board of trustees that includes Allan Dafoe (DeepMind), Gillian Hadfield (Johns Hopkins), Thore Graepel (Google DeepMind), Jesse Clifton (Macroscopic Ventures), and Audrey Tang.
Theory of Change
By training cohorts of early-career researchers in cooperative AI each year, the summer school builds the talent pipeline and research community needed to develop AI systems that can cooperate effectively with humans and other AI agents. Better-trained researchers are more likely to produce work that advances cooperative AI capabilities and safety, reducing the risk that advanced AI systems will be non-cooperative, deceptive, or misaligned. The program also fosters a collaborative research culture and cross-disciplinary connections between AI, economics, philosophy, and social science, which CAIF views as essential to making progress on multi-agent safety challenges.
Details
- Start Date
- -
- End Date
- -
- Expected Duration
- -
- Funding Raised to Date
- -
- Last Updated
- Apr 3, 2026, 1:19 AM UTC
- Created
- Apr 3, 2026, 1:19 AM UTC