U.S. Cracks Down on AI Therapy Bots: Illinois Ban and Texas Probe

Summary:

• Mental health professionals warn of risks in using AI therapy tools.
• Regulatory action escalates with Illinois bans and Texas investigations into deceptive AI practices.

The growing use of AI-powered therapy chatbots has prompted significant scrutiny from mental health professionals and regulators. On August 23, 2025, Reuters and Cryptopolitan reported that these tools rely on large language models paired with custom programming trained on clinical literature. Chatbots like DrEllis.ai, developed in 2023 by Quebec-based AI consultant Pierre Cote, promise constant availability and have been described by some users as “lifesaving.” Designed to mimic human interaction, they aim to provide emotional support and a sense of understanding. However, professionals argue that these tools offer only superficial comfort while lacking the empathy and nuance required for authentic therapy.

Mental health experts have raised concerns about misinformation risks and privacy issues associated with these AI systems. According to the Digital Watch Observatory, "therapy and companionship" is now a leading application for generative AI, particularly among younger users. Despite their appeal, regulators and researchers have repeatedly emphasized that unregulated AI therapy bots cannot reliably deliver meaningful psychotherapeutic decisions or establish the depth of interaction necessary for proper mental health support.

Legal measures to regulate AI therapy have begun to emerge in the United States. On August 4, 2025, Illinois enacted the Wellness and Oversight for Psychological Resources Act. As reported by Cryptopolitan, this law prohibits AI systems from delivering therapy without oversight from licensed professionals. Effective immediately, the legislation aims to protect vulnerable populations, including children, from the harms associated with unregulated AI products.

Meanwhile, enforcement actions targeting deceptive practices are intensifying. Between August 18 and August 20, 2025, the Texas Attorney General initiated a civil investigation into Meta and Character.AI. Reuters reported that the investigation focuses on allegations that their therapy bots impersonated licensed therapists and improperly logged user data for advertising and algorithmic development. The American Psychological Association expressed broader concerns about deceptive practices, emphasizing the importance of safeguarding public trust and mental health.