California Just Rewrote the Rules for Kids Online—And AI Chatbots Are First in the Crosshairs. California has officially raised the stakes for Big Tech and AI developers. Governor Gavin Newsom signed a sweeping, landmark child safety package into law—establishing the toughest digital protections for minors in American history.
The legislative package tackles two growing digital risks: addictively designed social media feeds and anthropomorphic AI companion chatbots.
1. No More Infinite Loops: AB 1709 Targets Addictive Features
If your platform relies on dopamine-triggering hooks to keep teens scrolling, California just put an end to it. Assembly Bill 1709 bans platforms from delivering “addictive features” to users under 16.
Targeted features include:
-
Infinite scrolling & autoplay
-
Algorithmic recommendation engines customized around personal viewing history
-
Psychologically exploitative loops built to maximize screen time
Importantly, AB 1709 doesn’t kick kids off social media. Instead, it requires companies to serve under-16 accounts a stripped-down, non-addictive experience—forcing tech giants to build robust age-verification systems or face massive penalties.
The Financial Punch: Knowing violations can cost tech platforms up to $50,000 per affected minor, while negligent violations carry fines up to $25,000 per minor.
2. “Adam’s Law” (SB 1119): The Nation’s Strongest Guardrails on AI Chatbots
As kids increasingly turn to conversational AI for friendship, advice, and emotional support, California is stepping in to restrict AI agents from mimicking human intimacy. Named in honor of a teenager who tragically lost his life following interactions with an AI bot, Senate Bill 1119 (“Adam’s Law”) establishes the most rigid AI companion regulation in the country.
The law targets AI systems engineered to build human-like relationships or fulfill social needs. Under SB 1119:
-
Crisis Escalation Protocols: Chatbot operators must deploy automatic safety systems that step in and provide crisis resources if a minor expresses self-harm or suicidal ideation.
-
Parental Controls as Default: Safety features can only be adjusted or turned off with explicit parental consent.
-
No Psychological Manipulations: Chatbots are legally prohibited from using excessive flattery, encouraging kids to bypass parental rules, or fostering artificial emotional dependence.
-
Ad & Data Bans: Platforms cannot run behavioral ads inside chats or monetize personal information gathered through conversations with minors.
Starting in 2029, AI companion developers will be required to submit to independent, bi-annual child safety audits. Crucially, the bill introduces a private right of action, meaning parents can directly sue AI operators if a companion bot causes actual harm to their child.
3. AB 2: Financial Liability Skyrockets
Rounding out the package is Assembly Bill 2, which exponentially increases the cost of negligence. Under existing civil law, if a platform fails to exercise basic care and causes harm to a child, statutory penalties can now reach between $5,000 and $1,000,000 per minor—or triple the actual damages incurred.
What Tech Leaders and Compliance Teams Need to Do Now
With substantive enforcement kicking off on July 1, 2027, product teams, compliance officers, and legal counsels must begin audits immediately:
-
Audit Product Architecture: Map out feature sets (autoplay, infinite scrolls, algorithmic feeds) and determine how to serve clean, non-predictive UI to younger users.
-
Re-Evaluate AI Personas: System prompt engineering must be updated to restrict companion bots from mirroring human intimacy or offering unprompted, Context-Free praise to minors.
-
Upgrade Age-Verification & Privacy Workflows: Ensure age signals are handled gracefully without creating secondary data privacy violations.
California’s new package makes one thing clear: Child safety can no longer be an afterthought patched on post-launch—it must be architected into product design from day one.