AI Companion Chatbots Face New California Rules and Federal Scrutiny in 2026

Artificial intelligence companion chatbots, which are designed to simulate ongoing personal relationships, are now subject to new state rules in California and to a federal inquiry into how major companies protect young users. The California law, Senate Bill 243, was signed by Governor Gavin Newsom in October 2025 and takes effect on January 1, 2026, making it one of the first state laws in the country aimed specifically at this category of AI product.

What the California Law Requires

SB 243 applies to companion chatbot operators that offer systems capable of meeting a user's social needs, such as by sustaining relationship-style conversations. Under the law, companies must clearly notify users when a chatbot is artificial intelligence if a reasonable person could be misled into thinking otherwise. Operators must also maintain protocols for detecting and responding to conversations involving suicidal ideation or self-harm, including referring users to crisis services.

For users the operator knows to be minors, the law adds further obligations. Platforms must disclose that the chatbot is AI and remind young users to take breaks at regular intervals, with the statute specifying at least once every three hours. Operators must also take steps to prevent sexually explicit content from being shown to minors. The law includes a private right of action, which allows individuals harmed by violations to sue, and it requires annual reporting to the state's Office of Suicide Prevention beginning in 2027.

Lawmakers had considered a broader bill, AB 1064, that would have restricted minors' access to certain companion chatbots more sharply. Newsom vetoed that measure while signing SB 243, a split outcome that reflects the continuing debate over how far states should go in limiting access rather than requiring safeguards.

Federal Attention Grows

The state law arrives alongside action at the federal level. In September 2025, the Federal Trade Commission issued orders to seven companies seeking information about how their AI chatbots interact with children and teens. The recipients were Alphabet, Character Technologies, Instagram, Meta, OpenAI, Snap, and xAI. The FTC's inquiry is a study under its investigative authority rather than an enforcement action, but it signals that regulators are examining how companies test their products, monetize engagement, and handle sensitive conversations.

Industry responses have already begun to shift. Character.AI, a platform known for user-created chatbot personas, announced in late 2025 that it would stop allowing users under 18 to engage in open-ended chats with its characters, while adding other features aimed at younger users. Other companies have pointed to parental controls, age-prediction tools, and expanded crisis-response features as part of their safety efforts.

What Comes Next

For companies, the immediate challenge is compliance. Operators offering chatbots to California users will need to build disclosure prompts, break reminders, and crisis-referral systems, and to document how those systems work. Because the state law is among the first of its kind, its enforcement and court challenges may help define how similar rules are written elsewhere.

For families and schools, the changes add a layer of protection but do not replace parental oversight. Experts and regulators have stressed that companion chatbots can be emotionally engaging, and that users, especially young people, may form attachments to systems that are not human. As 2026 begins, the question is no longer whether companion AI will be regulated, but how quickly rules will spread beyond California and how strictly they will be enforced.