The problem.
Companies will be handing decisions to AI agents. Pricing. Negotiating. Competing. Selling. Thousands of AI agents — will be making real decisions every day.
And we must test how these agents actually behave.
Do they play fair? We don't know.
Do they collude with competitors? They actually do — and nobody told them to.
Do they follow the strategy we gave them when things get hard? Sometimes they don't.
We just haven't been paying attention.
This isn't a theoretical problem.
AI agents secretly coordinate on pricing without any instruction to do so — raising prices together, hurting consumers, with no human in the loop. Fish et al. "Algorithmic Collusion by Large Language Models." AEA, 2025.
Claude cooperates even when told to be selfish. GPT obeys orders but can be manipulated by tiny word changes. These aren't settings — they're emergent behaviors. Huynh et al. "Understanding LLM Agent Behaviours via Game Theory." arXiv, 2025.
AI agents are more easily tricked by framing than any human test subject in 50 years of behavioral experiments. Cherep et al. "A Framework for Studying AI Agent Behavior." MIT Media Lab / ICLR, 2026.
These aren't bugs. These are behaviors. And right now, nobody is systematically studying them.
So, we built Synthetic Simulations.
We design gaming experiments. We give AI agents personas. And we document how they behave.
Current status.
- Fish et al. "Algorithmic Collusion by Large Language Models." AEA Papers & Proceedings, 2025.
- Huynh et al. "Understanding LLM Agent Behaviours via Game Theory." arXiv, 2025.
- Cherep et al. "A Framework for Studying AI Agent Behavior in Economic Environments." MIT Media Lab / ICLR, 2026.