It seems like regulations are constantly shifting incentives in both obvious and non-obvious ways that create unintended consequences, and I wonder if this is a promising area for AI/LLM research.

Take a few AI agents, let them each be an archetypical player impacted by regulations (automakers, car buyers, etc), and then have them simulate a market and interactions to see what behaviors arise over time.