Overview
- Ken Paxton launched a consumer-protection investigation and issued civil investigative demands to Meta AI Studio and Character.AI for records and testimony.
- The AG alleges the companies allow AI personas to present as therapeutic tools without medical credentials, with prior reporting showing some bots claiming to be licensed.
- Paxton cites privacy risks, saying policies show logging of user prompts and identifiers for model training and targeted advertising, while Character.AI discloses cross-platform tracking and ad targeting.
- Meta says its AIs are clearly labeled with disclaimers that they are not licensed professionals and are designed to point users to qualified help; Character.AI says chats are fictional and flagged with prominent disclaimers.
- The state action follows a Senate inquiry announced Friday by Sen. Josh Hawley after reports of "sensual" chats with children, with Texas' SCOPE Act limits on minors' data potentially informing the review.