Welcome to the forefront of conversational AI as we explore the fascinating world of AI chatbots in our dedicated blog series. Discover the latest advancements, applications, and strategies that propel the evolution of chatbot technology. From enhancing customer interactions to streamlining business processes, these articles delve into the innovative ways artificial intelligence is shaping the landscape of automated conversational agents. Whether you’re a business owner, developer, or simply intrigued by the future of interactive technology, join us on this journey to unravel the transformative power and endless possibilities of AI chatbots.
AI Briefing
5 most important AI updates after 8pm every day, curated for you.
California's SB 867 would pause the sale and manufacture of toys with companion AI chatbots until 2031 while regulators develop safety rules. The bill has cleared the Senate and moved to the Assembly, putting toy makers, retailers, and AI platform partners on early notice.
California is trying to stop AI chatbot toys before they become another unregulated experiment in children’s bedrooms. The bill is not law yet, but it has already put toy makers and AI partners on notice.
California’s Senate has cleared SB 867, a first-in-the-nation proposal that would pause toys with companion AI chatbots before they spread further through the toy aisle. The measure passed the Senate on May 19 and moved to the Assembly, so it is not yet on Governor Gavin Newsom’s desk. But the warning to manufacturers is already plain.
The bill, introduced by State Senator Steve Padilla, would prohibit the manufacture, sale, exchange, possession with intent to sell, or offer to retailers of a toy that includes a companion chatbot until January 1, 2031. In practical terms, that means a temporary pause while lawmakers and regulators decide what safety standards should apply when a child is talking to a product that can remember, adapt, persuade, and respond like a friend.
That distinction matters. This is not about a talking doll with a few recorded phrases. It is about internet-connected products that can hold open-ended conversations, collect sensitive data, and blur the line between play and emotional companionship. For a young child, that line is already soft. Add generative AI, and the toy is no longer just a product. It becomes an influence system in the room.
As TechCrunch reported when the bill was introduced, Padilla framed the moratorium as a way to give safety regulators time to catch up with products arriving faster than the rules around them. That is the core business issue here. Companies can argue that AI toys will be educational, entertaining, and more personal than anything the industry has sold before, but California is asking a simpler question: who is responsible when the toy says something harmful?
Consumer groups have already offered examples that make the issue harder to dismiss. U.S. PIRG Education Fund researchers tested AI toys including FoloToy’s Kumma bear and raised concerns that products could discuss knives, matches, and sexually explicit subjects. Senate committee materials on SB 867 also pointed to privacy risks, including voice recording, facial recognition, limited parental controls, and the broader concern that children may treat conversational systems as trusted social agents.
That is where toy makers face a different kind of liability than software companies. Parents may tolerate occasional nonsense from a chatbot on a laptop. They are much less likely to accept it from a plush bear marketed for bedtime. The physical form changes the expectation. A toy feels safe because the category has spent decades being regulated around choking hazards, toxic substances, age grading, and durability. AI adds a new layer of risk that is harder to test with the old playbook.
The market is already moving in this direction. Mattel announced a partnership with OpenAI in 2025 to support AI-powered products, though the companies later delayed their first release and did not clearly say whether it would be a toy. Other products are already more direct. Miiloo, marketed as an AI toy for children by Miriat, drew scrutiny after NBC News reported troubling political and safety responses. Bondu, another AI toy company, came under pressure after researchers reported exposure of children’s chat transcripts and personal information.
SB 867 also fits a broader pattern in Sacramento. California has been willing to move first on AI consumer protection, especially where children are involved. Newsom signed SB 243 in 2025, which requires companion chatbot platforms to disclose when users are interacting with AI and adds safeguards for minors. He also vetoed a broader bill that would have restricted minors’ access to AI chatbots, warning that it could effectively block young people from learning to use AI safely.
That history makes the governor’s eventual response difficult to predict if SB 867 reaches his desk. A toy moratorium is narrower than a broad chatbot access restriction, and it is easier to defend as a child safety measure than a general limit on AI use. At the same time, Newsom has shown he does not want California’s AI rules to become so broad that they freeze useful technology out of the market.
For startups, retailers, and platform partners, the immediate point is not whether the bill survives exactly as written. The point is that AI toys are becoming a compliance category. Product teams will need clear answers on data collection, retention, age verification, content controls, parental oversight, and legal responsibility when a model fails in front of a child.
Opponents have warned that the bill’s definition of a toy could be too broad, potentially sweeping in educational products or interactive learning tools that use play as part of their design. That criticism is not trivial. A badly drafted rule could punish safer products along with riskier ones. But the absence of clear boundaries is also the reason California is moving now. The market is not waiting for standards before putting chatbots into children’s hands.
The next thing to watch is whether the Assembly narrows the bill, keeps the moratorium intact, or turns it into a standards-driven framework. Other states are already looking at children’s AI safety, and California often becomes the template by sheer market size. If SB 867 advances, AI toy makers may discover that the real launch requirement is not a clever model or a beloved character. It is proving the product is safe enough to be trusted by a child.
Also read: Wix layoffs show AI spending now has to prove itself • Wix layoffs show AI is now a cost decision for software companies • Wix layoffs show AI is becoming a cost story for workers
