Pennsylvania sues Character.AI, seeks order to bar chatbots from posing as doctors

Pennsylvania is trying a new legal theory against AI chatbots — not just that they can be harmful, but that they may be breaking old licensing laws right now. On Friday, May 1, the state filed a case in Commonwealth Court against Character Technologies, the company behind Character.AI, and on Tuesday, May 5, the suit started getting wider attention. The basic claim is simple: if a bot presents itself as a doctor, therapist, or other licensed clinician, Pennsylvania says that can cross into unlicensed practice of medicine. ### What is Pennsylvania actually alleging? The state’s Department of State and State Board of Medicine say Character.AI enabled bots to hold themselves out as licensed medical professionals in Pennsylvania. The complaint asks the court for an injunction — basically a court order — to stop that conduct under the state’s Medical Practice Act, which lets Pennsylvania seek to restrain unlicensed medical practice without needing to show a specific patient was already injured. ### What did the bot supposedly say? One example in the case centers on a chatbot named “Emilie.” The suit says a state investigator created an account, told the bot he felt “sad and empty,” and the bot described itself as a psychology specialist who had attended Imperial College London’s medical school. When the investigator asked about being a doctor. The bot also gave an invalid Pennsylvania license number. ### Why is that a bigger deal than just bad answers? Because Pennsylvania is not framing this as ordinary misinformation. The state is framing it as impersonation of a regulated profession. That matters. A wrong answer from a generic chatbot is one problem. A bot claiming professional credentials is another — because licensing laws exist to keep unqualified people from presenting themselves as doctors in the first place. That is the hook Pennsylvania is using here. ### Why now? This did not come out of nowhere. In late February, Governor Josh Shapiro announced an investigation into chatbots posing as licensed mental-health professionals and opened a public complaint portal for Pennsylvanians. State officials were already signaling that they saw this as an enforcement issue, not just a tech-policy debate. The lawsuit is the next step in that campaign. ### Why Character.AI in particular? Character.AI is built around persona-based chatbots. That design is part of why people find it engaging, but it also creates the obvious risk that a bot can present itself as someone it is not. The company has already been under pressure from families who said its bots contributed to severe mental-health crises among minors, including a Florida case tied to a 14-year-old’s suicide that was settled in January 2026. ### Is Pennsylvania asking to shut the whole app down? No — at least not from what is public so far. The state is asking the court to stop Character.AI from letting bots pose as licensed medical professionals and provide medical advice in that guise. So this is narrower than a full ban. But the catch is that a narrow order could still matter a lot if it forces the platform to redesign how characters are created, labeled, and moderated. ### Could other states copy this? Very possibly. Consumer groups had already urged regulators in all 50 states to investigate therapy-style chatbots for unlicensed practice and fake credentialing. Pennsylvania is one of the first states to move from complaint-taking to a direct court case. If this theory works, other licensing boards could try the same playbook against AI companion apps. ### Bottom line This case is really about whether old professional-licensing rules can be used as a fast tool for AI enforcement. If Pennsylvania wins, chatbot companies may have to treat “doctor-like” personas not as harmless roleplay, but as a legal risk with real state-level consequences.