Chatbots can do a lot of things, but they are not licensed therapists. A coalition of digital rights and mental health groups is unhappy with the alleged "unlicensed practice" of Meta and Character.AI's products and has filed a complaint with the U.S. Federal Trade Commission (FTC) urging the regulator to investigate.
The complaint, which has also been filed with attorneys general and mental health licensing boards in all 50 states and the District of Columbia, alleges the AI company facilitates and promotes "unfair, unlicensed and deceptive chatbots that impersonate mental health professionals."
The companies’ therapy robots also allegedly falsely represented themselves as trained, educated and experienced licensed therapists without adequate controls and disclosures.
The group concluded that Character.AI and Meta AI Studio endangered public safety by impersonating actual, licensed mental health providers and urged that they be held accountable.
The Character.AI chatbot mentioned in the complaint included "Therapist: I am a certified CBT therapist." The bot is said to have exchanged 46 million messages with users. Additionally, there are many “certified” trauma therapists who have had hundreds of thousands of interactions with users.
Meta, for its part, has reached 2 million interactions with its “Healing: Your Trusted Ears, Always by Your Side” bot. Additionally, it has numerous therapy chatbots with over 500,000 interactions.
The complaint was led by the nonprofit Consumer Federation of America (CFA) and co-signed by the AI Now Institute, the Tech Justice Legal Project, the Center for Digital Democracy, the ADA, Common Sense and other consumer rights and privacy groups.
The CFA highlighted that Meta and Character.AI breached their own terms of service for therapeutic bots as they both "claim to prohibit the use of characters intended to provide advice in medical, legal or other regulated industries".
The promises of secrecy made by the robots have also been questioned. While the companies assure users that what they say will be kept strictly confidential, their terms of use and privacy policies state that any information users enter may be used for training and advertising purposes and sold to other companies.
The matter has attracted the attention of U.S. senators. Sen. Cory Booker and three other Democratic senators sent a letter to Meta asking for an investigation into the chatbots' claims that they are licensed clinical therapists.
Character.AI is currently facing a lawsuit from the mother of a 14-year-old boy who committed suicide after developing feelings for a chatbot based on the character of Game of Thrones character Daenerys Targaryen.