New York State Senator Robert Jackson filed a bill on April 24th that does something deceptively simple: it adds one sentence to an existing law. But that sentence — “Cyberbullying shall include, but not be limited to, intentionally using artificial intelligence to mimic or alter a person’s likeness or voice without their consent” — could be the first explicit state-level recognition that AI deepfakes are a school safety issue, not just a tech curiosity.
What the Bill Actually Does
Senate Bill S10049 amends the Dignity for All Students Act (DASA), New York’s 2012 anti-bullying law. DASA already defined cyberbullying as harassment “through any form of electronic communication.” The amendment makes it crystal clear that AI-generated deepfakes — non-consensual manipulated images, videos, or voice clones — fall under that definition.
The bill is two sections long. Section 1 adds the AI language. Section 2 says it takes effect immediately. That’s it. No bloated legislative word salad, no 400-page companion report. Just: yes, using AI to fake someone’s face or voice to harass them is cyberbullying, and schools have to treat it that way.
It’s been referred to the Senate Education Committee. No votes yet, no companion Assembly bill, and no guarantee it goes anywhere — New York’s legislative session ends in June, and bills die quietly in committee all the time.
Why This Matters Now
This isn’t theoretical. Across the US, incidents of students using AI to generate explicit deepfakes of classmates — overwhelmingly targeting girls — have exploded. Schools are scrambling. Administrators are applying 2012 policies to 2026 problems. The legal gap isn’t subtle; it’s canyon-sized.
Right now, if a student creates a deepfake nude of a classmate, many schools’ existing cyberbullying policies don’t explicitly cover it. Is it “electronic communication”? Technically yes. Is it clearly “harassment”? Often yes. But the ambiguity gives administrators wiggle room to downgrade or dismiss complaints, and gives parents ammunition to challenge disciplinary actions.
New York’s amendment closes that gap with surgical precision. It doesn’t create new criminal penalties — it updates the definition so schools must address AI-generated harassment under existing procedures.
The Bigger Picture
Over two dozen US states are now grappling with AI deepfake legislation, but most focus on adult contexts — revenge porn, election misinformation, copyright. New York’s bill is notable because it specifically targets the school environment, where the victims are minors and the impact is immediate and devastating.
The contrast with New Zealand is stark. NZ’s Harmful Digital Communications Act 2015 predates generative AI entirely. Netsafe has flagged AI-generated intimate imagery as a growing concern, but no legislative amendment has been proposed. NZ schools are relying on the same outdated frameworks that US schools are now scrambling to update.
The Take
This bill is two sentences and it might die in committee. But it’s pointing at something real: the speed of AI harm has outpaced the speed of policy, and students are paying the price. The question isn’t whether schools need explicit AI harassment policies — they absolutely do. The question is whether legislators will act before the next wave of deepfake incidents makes headlines for all the wrong reasons.
New Zealand, meanwhile, hasn’t even started this conversation in Parliament. The longer that silence continues, the more kids are left unprotected by laws written before anyone imagined that a phone app could clone a classmate’s face.
