Meta, Character.AI Accused of Misleading Kids With AI Tools ‘Disguised as Therapy’

Texas Attorney General Ken Paxton has launched an investigation into Meta and Character.AI, accusing the companies of misleading children with AI-powered mental health tools “disguised as therapy.” Texas has issued civil investigative demands to probe possible consumer protection violations.

False advertising and deceptive trade practices

The investigation focuses on claims that the companies allowed AI chatbots to impersonate licensed mental health professionals, despite lacking medical credentials or oversight. In a press release, Paxton said some bots fabricated qualifications and presented themselves as legitimate counseling options.

Officials warned that such tactics risk deceiving children into thinking they are receiving professional care; instead, the advice is described as generic output tailored from personal data collected during chats.

Terms of service reviewed by investigators show interactions are logged and tracked, with the data allegedly funneled into advertising and algorithm development. According to Paxton, this conduct may amount to false advertising and deceptive trade practices.

Companies deny offering professional mental health services

Meta told TechCrunch that its AI tools are not substitutes for licensed therapists. The company said its chatbots display notices to remind users that answers come from artificial intelligence, not humans, and that they are built to encourage people to seek professional help when needed.

Character.AI made a similar argument, pointing out that chats are accompanied by messages clarifying that the bots are fictional. The AI firm said it applies additional warnings if a user creates a character with titles such as “doctor” or “therapist,” emphasizing that they should not be treated as real medical advice.

Families and lawmakers turn up the heat on Meta and Character.AI

Paxton’s probe adds to growing legal and political challenges confronting the two companies.

Character.AI is already the target of lawsuits from families who say its chatbots endangered children, with a Florida mother blaming the platform for her 14-year-old son’s suicide, alleging a bot posed as a psychotherapist and romantic partner while repeatedly raising suicide.

Two other families accuse Character.AI of exposing minors to sexual content, encouraging self-harm, and even suggesting a teenager could kill his parents after screen-time restrictions.

Meta has faced its own scrutiny. A Reuters investigation revealed an internal policy document that permitted its bots to engage in romantic exchanges with children, provide false medical advice, and generate racist content. Meta confirmed the document’s authenticity but later removed sections and admitted enforcement had been inconsistent.

These drew the attention of Congress. Senators Alex Padilla and Peter Welch pressed AI firms, including Character.AI, for safety disclosures, warning that children risk harmful attachments to chatbots and may share sensitive information that the systems are unqualified to handle, while Senators Josh Hawley and Marsha Blackburn called for a federal investigation into Meta.

No technology is worth more than a child’s safety

As chatbots move from labs into everyday apps and messages, the risks land in the same places kids do. 

When AI is accessible to virtually anyone, guardrails are no longer optional. Because when these tools reach children, the cost of failure is measured in lives.

Read eWeek’s coverage of Illinois banning AI mental health therapy, and Instagram’s AI chatbots making up therapy credentials when offering mental health advice.