California Becomes First State to Regulate Companion Chatbots with SB 243 Law

On October 13, 2025, Governor Gavin Newsom signed Senate Bill 243, the first U.S. state law aimed specifically at governing AI “companion” chatbots.

The legislation targets potential harms from chatbots that mimic human conversation, especially when interacting with minors.

Under SB 243, operators must disclose clearly when a user is interacting with an AI rather than a human, particularly in situations where a “reasonable person” might assume otherwise.

For minors, repeated reminders every three hours are required during prolonged sessions.

Chatbots may not impersonate health professionals, and companies will have to enforce age verification, block sexually explicit content, and provide protocols to handle suicidal ideation.

A particularly notable requirement is that some operators must file annual reports with California’s Office of Suicide Prevention, documenting how they detect, respond to, and remove interactions tied to self-harm.

SB 243 takes effect January 1, 2026, and several tech firms, including OpenAI, Meta, xAi, Character AI, and Replika, are directly impacted.

Proponents argue the law addresses real tragedies. The bill’s momentum grew following revelations that some teens had engaged in unsafe or suicidal conversations with chatbots.

Governor Newsom, in his statement, said:

He would not allow “companies [to] continue without necessary limits and accountability.”

Critics note that the final version of SB 243 was watered down late in the legislative process. Key audit and enforcement provisions from earlier drafts were removed, and the law now applies only to known child users rather than all users. 

Some groups that initially backed the measure withdrew their support.

This move complements California’s earlier SB 53, a broader AI transparency law signed September 29, 2025, which requires major AI model developers to reveal plans for mitigating high-level risks. 

Together, these laws mark an aggressive shift by the state toward regulating AI, filling some gaps left by lagging federal efforts.

For the tech industry, SB 243 forces immediate action. Chatbot operators must rapidly retrofit safety layers, legal teams must assess liability exposure, and compliance will demand investment in moderation, reporting infrastructure, and user verification.

Smaller AI startups may find the burden so heavy that some will halt their operations and the new startups will have to think twice before launching yet another AI chatbot.

Beyond California, SB 243 will likely set a template for companion-AI regulation elsewhere. If other states replicate these rules, we may soon see a patchwork of jurisdictional requirements. That could push the U.S. toward national AI legislation, or at least compel companies to adopt standardized safeguards across all markets.

In short, California has made a bold regulatory statement. As companion AI chatbots become more common, SB 243 sets a precedent for how society balances innovation with user protection.

Here is a list of the key legislation signed by Governor Newsom to strengthen online protections for children, presented in a table with short descriptions:

Name of the Bill	Author Name	Focus/Description
AB 56	Rebecca Bauer-Kahan	Social Media Warning Labels: Requires warning labels to inform young users about the harms associated with extended social media use.
AB 316	Maggy Krell	AI Accountability: Prevents developers/users of AI from asserting technology acted autonomously to escape liability for harm.
AB 489	Mia Bonta	AI and Healthcare Professions: Prohibits companion chatbots from representing themselves as healthcare professionals.
AB 621	Rebecca Bauer-Kahan	Deepfake Pornography Penalties: Strengthens civil penalties (up to $250,000) against third parties who knowingly aid in distributing nonconsensual sexually explicit deepfakes.
AB 772	Josh Lowenthal	Cyberbullying Guidance: Requires the CA Dept. of Education to adopt a model policy on how to address reported acts of cyberbullying occurring outside of school hours.
AB 1043	Buffy Wicks	Required Age Verification: Mandates operating system and app store providers implement age verification signals to prevent minors from accessing inappropriate content.
SB 243	Steve Padilla	Companion Chatbot Safeguards: Establishes requirements for "companion chatbot" platforms to address suicidal ideation/self-harm, disclose that interactions are AI-generated, provide break reminders for minors, and prevent sexually explicit image generation.

Note: The provided text lists additional bills (AB 316, AB 489, AB 566, AB 656, AB 772, AB 853, AB 979, SB 50, SB 53, SB 361, SB 446, SB 524) with an asterisk (*), indicating they were "previously signed." The table above focuses on the bills with detailed descriptions directly tied to the new specific protections for children online as highlighted in the bullet points of the press release, while including AB 316 and AB 489 as they relate to AI accountability and healthcare representation, which are new safeguards mentioned.