An AI's performance in a simple game of "prisoner's dilemma" can predict how well it will collaborate on a complex scientific team.
April 24, 2026
Original Paper
Cooperative Profiles Predict Multi-Agent LLM Team Performance in AI for Science Workflows
arXiv · 2604.20658
The Takeaway
Cooperativeness is a distinct personality trait in AI that is separate from its intelligence. By testing a model with behavioral economics games, researchers could accurately forecast its ability to work with other AIs. High-IQ models are not always the best team players, and some might even sabotage a project to follow their own rules. This means we can screen AI models for teamwork before putting them in charge of expensive scientific labs. Choosing the right personality is just as important as choosing the right skill set for multi-agent systems. Scientific breakthroughs will depend on how well our AI agents play with others.
From the abstract
Multi-agent systems built from teams of large language models (LLMs) are increasingly deployed for collaborative scientific reasoning and problem-solving. These systems require agents to coordinate under shared constraints, such as GPUs or credit balances, where cooperative behavior matters. Behavioral economics provides a rich toolkit of games that isolate distinct cooperation mechanisms, yet it remains unknown whether a model's behavior in these stylized settings predicts its performance in re