US states flood legislatures with AI chatbot and algorithm bills

The latest TCAI update shows that US AI governance is being built bottom‑up in statehouses rather than waiting for one grand federal framework. We’re seeing dozens of convergent ideas: mandatory disclosures for AI chatbots, bans on sexually explicit companion bots for minors, guardrails on AI in health insurance and mental health, and scrutiny of algorithmic pricing in housing and consumer markets.​([transparencycoalition.ai](https://www.transparencycoalition.ai/news/ai-legislative-update-jan23-2026)) While the bills differ in scope, the pattern is clear – legislators are moving fastest where AI directly touches vulnerable populations or pocketbook issues.

From a race‑to‑AGI vantage point, this web of state rules won’t stop frontier research, but it will shape how and where powerful models can be deployed commercially. Labs and platforms will need compliance machinery capable of honoring dozens of overlapping constraints on features like agentic chat, image generation and automated decision‑making. That tends to favor larger players with legal budgets and in‑house policy teams, potentially entrenching current leaders. At the same time, poorly drafted state rules could create fragmentation risk: a model that is legal to deploy in one jurisdiction may need to be crippled or geofenced in another, complicating global rollouts of advanced assistants and agents.