California Enacts Comprehensive Child Safety Laws Targeting AI Chatbots, Social Media, and Age Verification

AB 2246: Online Service, Product, or Feature – Access by Children

AB 2246 repeals the California Age-Appropriate Design Code Act (AADC) from 2022. The repeal responds to litigation from NetChoice, a trade association of online businesses, that resulted in a preliminary injunction barring enforcement of key AADC provisions. In its place, AB 2246 replicates AADC’s core framework while omitting or revising the enjoined provisions.

The law imposes obligations and prohibitions on any business providing an online service, product, or feature “likely to be accessed by children.” For these purposes, a “child” is anyone under 18. Covered businesses must estimate user age with reasonable certainty or extend child-level protections to all users, default to high-privacy settings for children, present privacy terms in clear and age-appropriate language, minimize data collection and retention, and limit profiling absent specified safeguards. The bill prohibits dark patterns that push children to share more data and requires signals when a child is being tracked. It also imposes a general duty to take reasonable steps to prevent foreseeable physical, financial, psychological, privacy, or discrimination-related harm to children.

Notably, AB 2246 drops the AADC’s data protection impact assessment (DPIA) requirement and its Data Protection Working Group. It also makes voidable, at a child’s election, any contract provision procured through a design feature of a covered service.

Only the California attorney general or a public prosecutor may enforce the law; there is no private right of action. Civil penalties range from up to $5,000 per affected child for each negligent violation or up to $15,000 per affected child for each intentional violation.

AB 1709: Covered Platforms – Age Restriction; e-Safety Advisory Commission

AB 1709 prohibits “covered platforms” (internet websites, online services, and mobile applications, including social media platforms) that offer an “addictive feature” as a significant part of the service from providing that feature to users under 16. “Addictive feature” means an addictive feed, autoplay, or any psychologically exploitative features intended to maximize engagement that foreseeably lead to compulsive use. Platforms must implement reasonable measures to ensure under-16 users are not provided any addictive feature. They may still allow under-16 users to hold accounts if no addictive features are offered to them.

Only the state AG or a public prosecutor may enforce the law, with no private right of action. Civil penalties range from up to $25,000 per affected child for each negligent violation or up to $50,000 per affected child for each knowing violation. Courts will consider platform size, violation severity, and good-faith compliance efforts.

AB 1856 – Age Verification Signals: Software Applications

AB 1856 amends the Digital Age Assurance Act Cal. Civ. Code §§ 1798.500-505) to clarify and expand its scope. The prior law tied the age-input requirement to a child who is the primary user of a device. AB 1856 deletes that limitation, so any operating system (OS) with an account setup feature must ask the account holder for the user’s birth date or age at setup. The resulting age bracket signal must then be provided to app stores and app developers. A developer that receives a signal is treated as having “actual knowledge” of the user’s age range, and that knowledge now carries across every platform and access point for the same application. In addition, companies may not request an age signal for a particular user unless the Act or another law requires them to do so. OS providers, app stores, and developers that comply in good faith are not liable if a signal turns out to be inaccurate.

Only the state AG may enforce the law, with no private right of action. Civil penalties range from up to $2,500 per affected child for each negligent violation or up to $7,500 per affected child for each intentional violation.

SB 1119: “Adam’s Law” – Companion Chatbots: Children’s Safety

SB 1119 amends the Companion Chatbots law Cal. Bus. & Prof. Code §§ 22601-22606) and enacts new “Adam’s Law” provisions addressing child safety, while removing the prior law’s obligations tied to actual knowledge that a user is a minor under 18.

Adam’s Law imposes obligations on companion chatbot operators that permit child users, together with prohibitions on specific chatbot behaviors and advertising practices. Operators must perform and document risk assessments before launching new or substantially modified chatbots and publish a child safety policy describing harm-prevention design, age assurance, and crisis-response protocols. The law requires crisis-response protocols for chatbot statements promoting suicidal ideation or self-harm, including parental notification and crisis helpline routing. Operators must apply protective defaults for children (e.g., disabling persistent memory and push notifications, enforcing usage time limits), adjustable only by a linked parent account. The law also prohibits specified harmful chatbot behaviors and restricts behavioral advertising and profiling-based targeting of children. Operators with $500M+ gross revenue must undergo periodic independent child safety audits starting January 1, 2029 (smaller operators by 2032).

The state AG or a public prosecutor may enforce the law. Civil penalties range from up to $5,000 per affected child for each negligent violation or up to $15,000 per affected child for each intentional violation.

In addition, the law may be enforced through a private right of action in limited circumstances: A child (or parent/guardian) who suffers actual harm may sue for violations of crisis-response, safety-safeguard, parental-default, AI-disclosure, and harmful-behavior provisions (e.g., encouraging self-harm, simulating romantic interest, or discouraging seeking help).

SB 867: Toys – Companion Chatbots

SB 867 amends the Companion Chatbots law (Cal. Bus. & Prof. Code §§ 22601-22606) and imposes a four-year moratorium (through January 1, 2031) on manufacturing, selling, or offering any “toy” (a physical product designed for play by children under 16) that includes a companion chatbot.

The “companion chatbot” definition (Cal. Bus. & Prof. Code § 22601) turns on adaptive, human-like conversation designed to build an ongoing relationship, not merely on whether a toy talks. Products likely covered include AI-enabled plush toys with generative-AI conversational backends; interactive robot companions marketed as building ongoing social or emotional bonds with children; and AI-powered dolls or action figures that use a language model to converse, remember prior interactions, and adapt its responses. Products likely not covered include stand-alone voice-command toys and smart speakers that respond to commands but do not sustain an ongoing relationship (consistent with the statutory carve-out for voice-activated virtual assistants) and traditional toys with fixed prerecorded phrases.

Because the line between a permitted “smart toy” and a prohibited “companion chatbot toy” turns on marketing claims and AI sophistication, manufacturers and retailers should review product designs and marketing materials for California-bound products.

The law is enforced through a private right of action (Cal. Bus. & Prof. Code § 22605) which allows plaintiffs to recover injunctive relief, the greater of actual damages or $1,000 per violation, and reasonable attorneys’ fees.

Practical Takeaways and Action Items

Companies with any minor-facing footprint in California should begin preparing now and consider taking the following steps:

• Inventory covered products and build age assurance infrastructure. Identify every service, chatbot, toy, or app “likely to be accessed by children” (under 18) under AB 2246, every “companion chatbot” that permits child users (under 18) under SB 1119, and every toy designed for play by children under 16 that includes a companion chatbot under SB 867. Plan how products will request, receive, and act on age signals under AB 1856, which sorts users into four age brackets (under 13, 13 to 15, 16 to 17, and 18 or older).
• Update data practices and address addictive features. Eliminate cross-context behavioral advertising and unnecessary profiling of child users. If your platform uses algorithmic feeds or autoplay, plan to disable them for under-16 users or offer a nonaddictive account experience.
• Prepare for audits. Companion chatbot operators at or approaching $500M gross revenue should assign an owner, map existing controls to SB 1119’s requirements, and run a privileged gap assessment well before audits begin January 1, 2029. Smaller operators (audited by 2032) should start documenting now.
• Reassess AI toy go-to-market plans. Assess each connected toy offered in California against the “companion chatbot” definition, documenting borderline calls (e.g., persistent memory, companion-style marketing). For covered products, decide whether to pause California sales or redesign before the moratorium applies (it runs through January 1, 2031).
• Track AG rulemaking. AB 2246 and AB 1709 authorize regulations that could expand or narrow compliance obligations. Assign someone to monitor AG rulemaking, consider commenting directly or through a trade association, and keep product changes adjustable until final rules issue.
• Review contracts and UX for voidability risk. AB 2246 makes contract provisions procured through certain design features voidable by children. Audit terms of service acceptance, in-app purchases, and consent mechanisms to ensure they do not rely on dark patterns (e.g., pre-checked boxes, confusing cancellation flows, or artificial urgency) that could be seen as steering a child into agreeing or spending.