California Companion Chatbot Toy Ban
California SB 867 — Toys: Companion Chatbots
United States • California
RAI-US-CA-SB86700-2026SB 867
California Companion Chatbot Toy Ban is Awaiting Entry in United States as of 20 Oct 2026, according to leginfo.legislature.ca.gov.
ActRisk ManagementMarket SurveillanceLiability and RedressCalifornia temporarily bans manufacture, sale, exchange, and specified commercial handling of companion chatbot toys for children under 16.
Summary
California SB 867 temporarily bans specified commercial activities involving physical toys for children under 16 that include companion chatbots. Approved September 10, 2026, it takes effect January 1, 2027; the prohibition sunsets January 1, 2031. The enacted sections do not specify penalty amounts or detailed enforcement procedures.
Full article
Read full text ↗Overview
California Senate Bill 867, officially titled “SB 867. Toys: companion chatbots.” and cited as Chapter 189, Statutes of 2026, establishes a temporary prohibition on specified commercial activities involving toys that include companion chatbots. Enacted during the 2025–2026 legislative session, it received Governor approval and was filed with the Secretary of State on September 10, 2026. Section 1 amends Business and Professions Code Section 22601; Section 2 adds and repeals Section 22604.5. The chaptered legislation establishes the operative definitions, prohibition, and sunset.
As a non-urgency statute, the act takes effect January 1, 2027, rather than on its approval date. Its status on October 20, 2026, is Awaiting Entry. The prohibition concerns physical products designed, marketed, or manufactured for play by children under 16 that include a companion chatbot; it does not cover every toy using artificial intelligence. Prohibited conduct includes manufacture, sale, exchange, possession with intent to sell or exchange, and specified retailer-facing exposure or offers. Simple possession or use by children or families is not expressly prohibited. Section 22604.5 is repealed January 1, 2031, but that sunset does not expressly repeal the Section 22601 amendments.
Definitions
Section 22601(a) defines artificial intelligence as an engineered or machine-based system that varies in autonomy and can, for explicit or implicit objectives, infer from received input how to generate outputs influencing physical or virtual environments. Section 22601(b)(1) defines a companion chatbot as an AI system with a natural language interface providing adaptive, human-like responses to user inputs and capable of meeting a user's social needs, including through anthropomorphic features and sustaining a relationship across multiple interactions. Section 22601(g) separately defines a toy as a physical product designed, marketed, or manufactured for use in play by children under 16 years of age.
Section 22601(b)(2) excludes bots used only for listed customer-service, business-operational, productivity and source-information analysis, internal-research, or technical-assistance purposes. It also excludes qualifying video-game bots restricted to game-related replies and unable to discuss mental health, self-harm, sexually explicit conduct, or sustain unrelated dialogue. Another exclusion covers qualifying stand-alone speaker and voice-command devices acting as voice-activated virtual assistants without sustained relationships or outputs likely to elicit emotional responses. These exclusions depend on their statutory conditions, not simply a product's label.
A companion chatbot platform is a platform allowing users to engage with companion chatbots. An operator is a person making such a platform available to a user in the state. “Office” means the Office of Suicide Prevention established under Health and Safety Code Section 131300. “Sexually explicit conduct” incorporates the meaning in 18 U.S.C. Section 2256. A video game is a game played on an electronic amusement device using a computer, microprocessor, or similar circuitry and its own monitor, or designed for use with a television or computer monitor, that interacts with the device's user.
Governance and Institutional Framework
The act operates through California's Business and Professions Code rather than establishing a separate licensing authority or AI product regulator. Section 22604.5 begins with “No person,” so the prohibition is not confined to companion chatbot platform operators. Manufacturers, sellers, and other commercial participants must assess whether their activities fall within its terms. The act creates no registration process, approval board, certification mechanism, or administrative permission to market otherwise prohibited toys. Application of the prohibition does not depend on an agency first classifying a particular product.
The definition of the Office of Suicide Prevention does not assign it new duties under Section 22604.5. Neither enacted section directs the office to inspect toys, grant exemptions, evaluate chatbot systems, or issue implementation regulations. The operator definition concerns platform availability to users in California, whereas the toy prohibition uses the broader term “person.” These provisions should be distinguished from the existing operator obligations described in the Legislative Counsel's Digest. The act does not establish a dedicated enforcement agency or toy-approval regime.
Key Focus Areas
The central focus is restricting commercial availability of physical children's play products incorporating socially responsive AI companions. Coverage requires both a qualifying toy and a qualifying companion chatbot. The toy definition uses alternative criteria: design, marketing, or manufacture for play by children under 16 can bring a product within scope. Classification therefore should not depend solely on a stated customer age or a manufacturer's chosen label. A purely digital service does not satisfy the physical-product requirement merely because children use it, although other chatbot provisions may apply independently.
The act uses a prohibition rather than a safety-certification pathway. Section 22604.5 provides no exception based on parental consent, warnings, content filters, testing results, or educational-benefit claims. These measures do not independently authorize conduct involving a product that remains within the prohibited category. Conversely, the statutory chatbot exclusions must be considered before classifying every conversational device as prohibited. The act introduces no new consent requirements, data-retention limits, model-evaluation benchmarks, or cybersecurity standards. Its principal intervention is a temporary restriction on commercial conduct, with definitions delimiting the affected products.
Implementation Framework
Practical implementation begins with product-by-product examination of physical form, intended play use, marketing, and conversational functionality. Businesses should assess whether a product satisfies Section 22601(g), whether its AI interface meets Section 22601(b)(1), and whether an exclusion under Section 22601(b)(2) applies. Specifications and actual functionality can inform that assessment, but the act does not prescribe a statutory audit or filing requirement. For covered products, businesses must avoid the conduct prohibited by Section 22604.5 beginning January 1, 2027. No additional transition period follows that effective date.
Commercial planning should address manufacturing, sales channels, exchanges, inventory held with intent to sell or exchange, and retailer-facing displays or offers. The chaptered text contains no express grandfather clause for products manufactured or acquired before commencement. Existing inventory must therefore be assessed against the prohibited activities rather than assumed eligible for later sale. Removing listings or changing functionality may be operational responses, but the act prescribes no particular remediation process. It does not expressly require destruction, consumer returns, or recall of products already in private possession.
Monitoring and Evaluation
The act creates no recurring reporting, inspection, incident-notification, independent audit, or formal evaluation duties. It sets no product-reassessment frequency and requires no published safety metrics. Businesses may nevertheless monitor firmware, model, service, and marketing changes because those changes can affect statutory classification. A device previously limited to technical assistance or game-related dialogue may require renewed assessment if its capabilities expand. Such monitoring is a practical implication of continued compliance, not a separately enacted monitoring program or documentation obligation.
The January 1, 2031, sunset is a definite repeal date for Section 22604.5, not a condition dependent on a study or replacement safety standards. The act requires no government report before repeal, specifies no measurable policy outcomes, and creates no automatic renewal mechanism. Later legislation could alter the sunset or introduce successor rules. Until then, the express repeal date should be distinguished from expectations about future regulation. The amendments to Section 22601 are not expressly subject to the same sunset.
Penalties, Liability, and Appeals
The Legislative Counsel's Digest describes violations of the toy prohibition as subject to the same civil liability as violations of existing companion chatbot provisions. The two enacted sections, however, specify no monetary penalty, minimum damages, damages cap, claimant eligibility, limitations period, or fee-shifting rule. They also contain no express civil-liability cross-reference establishing those details. A complete enforcement assessment requires examination of applicable surrounding code provisions. No particular fine or damages amount can be assigned to Section 22604.5 from this act alone.
The digest also describes misdemeanor treatment of existing restrictions on toys contaminated with toxic substances. That does not establish that SB 867 expressly makes companion chatbot toy violations misdemeanors: Section 22604.5 contains no criminal penalty language. The act provides no dedicated administrative appeal, reconsideration procedure, adjudicative body, or enforcement deadline. It does not specify whether particular remedies require proof of injury or another element. Those questions depend on applicable enforcement provisions and procedural law, not inference from penalties governing different toy hazards.
Relationship to Other Instruments
SB 867 complements rather than replaces the existing companion chatbot operator framework described in its digest. That framework requires clear and conspicuous notification that a chatbot is artificially generated and not human when a reasonable person would otherwise be misled into believing the interaction is with a human. It also includes obligations for users known to be minors, including disclosure that they are interacting with artificial intelligence. These are existing requirements, not new disclosure duties enacted by Section 22604.5. Compliance with them does not create an exception to the toy prohibition.
The amendment structure separates the continuing definitions in Section 22601 from the temporary prohibition in Section 22604.5. The under-16 threshold belongs to the toy definition and should not automatically replace age concepts in other chatbot provisions. The reference to Health and Safety Code Section 131300 identifies the Office of Suicide Prevention; the reference to 18 U.S.C. Section 2256 supplies a definition of sexually explicit conduct. Neither creates a new toy-approval regime. The act does not announce a replacement for existing restrictions concerning toxic substances in toys.
National and Federal Alignment
SB 867 is a California statute, not a federal prohibition on companion chatbot toys. It does not amend federal law, create a nationwide approval system, or state that federal product-safety compliance satisfies California's restriction. Its direct federal incorporation is Section 22601(f), adopting the meaning of sexually explicit conduct from Section 2256 of Title 18 of the United States Code. That definition matters to the video-game chatbot exclusion. This limited incorporation does not bring every federal rule concerning sexual content into the act.
The act contains no express federal preemption clause, detailed state-federal allocation of responsibilities, express nationwide-sales ban, special import provision, or comprehensive cross-border transaction rule. Businesses should assess federal requirements and California's prohibition separately, without assuming compliance with one resolves the other. The act does not settle how courts would resolve particular preemption or territorial-jurisdiction disputes. Such conclusions require additional authorities and transaction-specific facts, particularly where manufacturing, online offerings, fulfillment, or purchasers span multiple jurisdictions.
Implementation Timeline
| Milestone | Date | Notes |
|---|---|---|
| Governor approval | 2026-09-10 | The Governor approved SB 867. |
| Secretary of State filing and chaptering | 2026-09-10 | Filed as Chapter 189, Statutes of 2026, during the 2025–2026 legislative session. |
| Status assessment | 2026-10-20 | Awaiting Entry; approval did not make this non-urgency statute immediately effective. |
| Effective date | 2027-01-01 | The amended definitions and prohibition take effect under California's constitutional non-urgency effective-date rule. |
| Repeal of Section 22604.5 | 2031-01-01 | The prohibition section sunsets; this clause does not expressly repeal the Section 22601 amendments. |
Compliance Checklist
| Check | Required Action |
|---|---|
| Physical toy coverage | Assess whether the product is physical and designed, marketed, or manufactured for play by children under 16. This is a practical compliance assessment, not a prescribed filing. |
| Companion chatbot coverage | Assess its natural language interface, adaptive human-like responses, and capacity to meet social needs under Section 22601(b)(1). |
| Functional exclusions | Check the exact conditions for listed operational bots, restricted video-game bots, and qualifying stand-alone voice assistants; product labels alone are insufficient. |
| Manufacture, sale, and exchange | From January 1, 2027, do not manufacture, sell, or exchange a toy that includes a companion chatbot. |
| Commercial inventory | Do not possess a covered toy with intent to sell or exchange. Stock predating commencement has no express exemption solely on that basis. |
| Retailer-facing exposure and offers | Do not expose or offer a covered toy for sale or exchange to any retailer. |
| No safety-based authorization | Do not rely on parental consent, disclosures, filters, or testing as permission for prohibited conduct; the act supplies no such exception. |
| Separate operator obligations | Assess independently applicable companion chatbot operator requirements. Compliance with them does not override Section 22604.5. |
| Liability review | Consult surrounding code provisions before determining available claims or remedies; SB 867 states no monetary amounts or dedicated appeal process. |
| Sunset planning | Track January 1, 2031, and subsequent legislation. Repeal of Section 22604.5 does not itself repeal all related chatbot requirements. |
Sources and References
| Source | Type |
|---|---|
| California Legislative Information — SB 867 Chaptered Text, Chapter 189, Statutes of 2026 | official |
| California Legislative Information — SB 867 Official Bill Record | official |
| California Legislative Information — SB 867 Bill History | official |
| California Legislative Information — SB 867 Version Comparison | official |
More AI regulation in United States
AI regulation in United States: full overview
- California SB 947
- California SB 951 AI Layoff Notice
- United States - California - Surveillance Technology Regulation
- California AI Regulation Summary
- United States - Colorado - Synthetic Media Disclosure (HB 24-1147)
- United States - Colorado - AI in Insurance (SB 21-169)
- United States - Colorado - AI Consumer Protections (SB24-205)
- Colorado AI in Healthcare Act
© Regulations.AI — created on 20 Oct 2026 using Gemini 3.6 Flash · reviewed against official sources on 20 Oct 2026