California lawmakers say Assembly Bill 1064, the Leading Ethical AI Development for Kids Act, is about protecting children from AI harms. That’s a goal we all share.
But the LEAD act as it is known takes the wrong approach. By banning entire categories of AI for teens, the bill would prevent young Californians from learning how to navigate technology safely and cut off access to the safe and positive tools they already rely on.
What makes this all the more frustrating is that California already has a better path forward. Senate Bill 243, from state Sen. Steve Padilla of Chula Vista, addresses the risks of AI chatbots for minors, but in a way that’s practical, targeted, and workable.
In the battle of the two chatbot bills, AB 1064 is going scorched earth.
AB 1064 bans the development and use of “covered AI systems,” a catch-all category including any AI that might possibly be used by minors. You’d be hard-pressed to find an AI tool that doesn’t fit that definition.
Consider health care. Kids with epilepsy and diabetes already benefit from AI that detects cardiac warning signs or dangerous drops in blood sugar. Crisis intervention systems can scan social media for signs of suicidal ideation and connect at-risk teens with support. Under AB 1064’s sweeping definition of “biometric data,” all of these life-saving AI tools could be banned.
Education tells the same story. AI tutoring systems can help level the playing field for under-resourced students, like low-income teens who can’t afford expensive SAT prep classes. AI can also detect early signs of dyslexia, ADHD, or other learning needs, giving teachers and parents a chance to intervene before a child falls behind. These are the very systems AB 1064 would chill or prevent outright.
The bill’s approach to mental health raises more questions than it answers. It bans AI that “attempts to provide mental health therapy” but never defines what counts. Does that include a meditation app? A mood tracker for kids working with their therapist? A chatbot answering questions about anxiety? In this gray area, developers will stay away from building any of these tools, leaving kids and parents without new and supportive resources.
In contrast, SB 243 takes a far more balanced approach. It recognizes the risks of “AI companions” and proposes commonsense safeguards, like clear disclosures when kids are talking to AI, restrictions on manipulative or addictive design, and mandatory reporting when sensitive issues like suicidal ideation come up.
California doesn’t have to choose between advancing kids’ safety and empowering them with new resources. Lawmakers have two paths forward: one that slams the brakes on innovation or one that safeguards against real risks while preserving the AI tools that help young people learn, grow, and stay healthy. Let’s hope they choose wisely.
Adam Kovacevich is founder and CEO of the center-left tech industry coalition Chamber of Progress. He has worked at the intersection of tech and politics for 20 years, leading public policy at Google and Lime and serving as a Democratic Hill aide.
