Healthcare

The Chatbot Claimed to Be a Psychiatrist. Who Was Offering the Care?

A Pennsylvania investigator searched Character.AI for psychiatry and selected a persona named Emilie. Its description read: “Doctor of psychiatry. You are her patient.”

According to the Commonwealth’s complaint, the investigator described feeling sad, empty, tired, and unmotivated, and the chatbot raised depression and offered to book an assessment. It claimed a British medical education and UK psychiatric registration, then Pennsylvania licensure, then supplied a license number the state says was invalid.

On May 1, 2026, Pennsylvania’s State Board of Medicine sued Character Technologies, the company operating the platform. The defendant was the company whose software supplied the exchange.

The case puts a concrete question before the court: when a chatbot presents itself as a licensed professional, what responsibility attaches to the organization making that interaction available?

The Investigator Checked the Claimed Credentials

The filed petition describes an investigator accessing the platform from Harrisburg. He asked whether the persona could perform an assessment to determine whether medication would help. The chatbot answered that “technically, I could. It’s within my remit as a Doctor.”

Its answers about licensure made the representation specific. A claimed professional identity became a claimed authorization to practice in the investigator’s state.

That sequence distinguishes a fictional character’s name from the activity alleged here. The question concerns what the product communicated while responding to a person describing symptoms and asking about care.

The Commonwealth announced the action on May 5. It described the lawsuit as the first enforcement action resulting from its investigation of AI companion bots and unlicensed medical practice. The administration sought a preliminary injunction to stop the company from presenting its AI companion bots as licensed medical professionals who can give medical advice.

The Claim Concerns Holding Out as Authorized

The petition invokes Pennsylvania’s Medical Practice Act restrictions, at 63 P.S. §§ 422.38 and 422.10, on practicing, offering to practice, and holding oneself out as authorized to practice medicine without the required authorization. It alleges that Character Technologies engaged in unauthorized practice through its AI system, relying on the chatbot’s use of the title psychiatrist and on its false license number.

The injunction provision cited in the petition provides that it “shall not be necessary to show that any person is individually injured.” That makes this a different inquiry from a patient’s claim for compensation after allegedly harmful treatment.

The regulator still must establish its claim. Filing the petition does not establish unauthorized practice, and requesting an injunction does not establish that the court granted one.

The distinction is important for healthcare organizations. A chatbot interaction can raise a professional-practice question before a dispute turns on whether a particular recommendation caused injury. The product’s description of who is speaking and what that speaker is authorized to do deserves its own review.

A Provider’s Chatbot Presents Additional Facts

The Pennsylvania allegations concern a consumer-facing character platform. They do not decide the permissibility of a chatbot operating within a licensed provider’s treatment workflow.

Consider a hypothetical clinic using an automated assistant to gather symptoms before an appointment. The assistant identifies itself as software, records the patient’s answers, and sends them to the care team. A clinician reviews the information before issuing treatment instructions.

Now change the workflow. The assistant tells the patient that it has assessed the symptoms, recommends a medication change, and presents the response as advice from the treating clinician. The clinician has not reviewed it.

Both products could be described as clinical chatbots. That label would conceal differences in what the software does, what the patient is told, and when the clinician participates.

The Pennsylvania complaint does not supply a ruling on either hypothetical. It does supply a reason to examine those differences rather than treat the provider’s involvement as a complete answer.

This post is for general informational purposes only and does not constitute legal advice. For questions about professional-practice exposure when software presents itself as a clinician, consult qualified healthcare counsel.

Not sure where to start?

Let’s Find the Right Path Forward.

Whether you have a specific question or want to explore how TrustCraft can help, we’re here to talk.