44 AGs Warn Congress: KIDS Act Is a Liability Shield for AI Chatbots

TL;DR

• A bipartisan coalition of 44 state attorneys general — led by Minnesota AG Keith Ellison — sent Congress a letter opposing the federal KIDS Act (H.R. 7757), warning it would broadly preempt state laws regulating online harms to minors.
• The AGs explicitly called out AI chatbots, social media platforms, and online games as vectors of harm the bill would insulate from state-level enforcement.
• The letter signals mounting state resistance to federal legislation seen as prioritizing platform immunity over child safety, especially as AI recommendation systems and chatbots increasingly mediate kids’ digital lives.
• Supporters of the bill argue it creates consistent national standards, but state regulators accuse it of gutting their ability to police emerging AI-driven risks.

44 States Draw a Line on Federal AI Preemption

On May 27, 2026, a bipartisan group of 44 state attorneys general fired off a formal letter to Congress opposing the Kids Internet and Digital Safety Act — better known as H.R. 7757. Minnesota AG Keith Ellison co-led the coalition, which warned that the federal bill would “broadly preempt state laws addressing online harms to minors, including social media harms, obscenity, social gaming platforms, and artificial intelligence chatbots.”

That’s not gentle pushback. That’s a coordinated, cross-party shot across the bow aimed squarely at Congress and the tech industry.

The coalition doesn’t mince words: the KIDS Act would strip states of their power to regulate how platforms — including AI chatbot providers — expose kids to algorithmic manipulation, addictive design, and harmful content. It’s the kind of preemption fight that usually plays out in wonky legal briefs, but this one landed in the open with 44 signatures attached.

Why State AGs Think the KIDS Act Is a Trojan Horse for Big Tech

Here’s where it gets interesting. The AGs aren’t just worried about social media feeds or gaming loot boxes — they specifically name AI chatbots as a threat the bill would shield from state enforcement. That’s a big deal, because AI-driven conversational agents are exploding in popularity among minors, and we’re only beginning to understand the risks.

Chatbots don’t just answer questions anymore. They shape opinions, simulate relationships, and feed kids content based on opaque recommendation algorithms. Some are designed to maximize engagement — which, in plain English, means keeping users hooked. And when those users are 12 years old, the stakes aren’t abstract.

But the KIDS Act, according to this coalition, would yank enforcement power away from state attorneys general and hand it to a federal framework that — in their view — prioritizes platform immunity over accountability. It’s like handing the keys to the henhouse to the foxes and calling it a national security protocol.

I’ve covered tech policy long enough to know that when 44 state AGs agree on anything, the bill in question has a serious problem. This isn’t partisan theater. This is states — red and blue — saying they don’t trust Congress to protect their kids better than they can.

The coalition’s argument boils down to this: federal preemption sounds efficient until you realize it freezes out the enforcers closest to the ground. States like California, New York, and Utah have already passed or proposed laws targeting addictive design, algorithmic amplification, and AI-driven content feeds aimed at minors. If the KIDS Act passes as written, those laws could be nullified overnight.

And here’s the kicker — the bill’s supporters claim it establishes consistent national standards for online child safety. That sounds reasonable until you ask: consistent for whom? If the standard is set by a Congress that’s historically struggled to regulate tech companies with any teeth, states have every reason to be skeptical.

The tension here isn’t just legal. It’s philosophical. Do you trust a single federal rule to keep pace with the breakneck evolution of AI and social platforms? Or do you bet on 50 state laboratories experimenting, adapting, and enforcing in real time?

The AGs are clearly betting on the latter. They see the KIDS Act as a legislative straitjacket — one that locks in protections that might’ve made sense in 2023 but will be obsolete by 2027. AI chatbots are evolving faster than Congress can hold hearings. State enforcers want the flexibility to move when new risks emerge, not wait for a federal agency to catch up.

The AI Chatbot Problem Isn’t Theoretical Anymore

Let’s zoom out. AI chatbots aren’t fringe products anymore. They’re embedded in homework helpers, mental health apps, gaming platforms, and social networks. Kids interact with them daily, often without understanding they’re talking to a machine optimized for engagement, not well-being.

Some chatbots are genuinely useful. Others are designed to maximize session length, collect behavioral data, and push kids deeper into platform ecosystems. The line between helpful and exploitative is blurry, and it’s moving fast.

States have started to act. California’s Age-Appropriate Design Code — currently tied up in court — targets recommendation algorithms and data practices that harm minors. New York is debating similar measures. Utah passed a law requiring parental consent for social media accounts. These aren’t uniform, and they’re messy, but they’re responsive.

The KIDS Act would reportedly override much of that patchwork. It would set a federal floor — and, crucially, a federal ceiling — on what states can do. That might sound like clarity, but it also means states couldn’t go further even if local conditions demanded it.

And here’s where the competitive context matters. The EU and UK are already setting strict rules for recommender systems, synthetic media, and AI-driven content aimed at minors. European regulators are moving faster and more aggressively than their US counterparts. If the KIDS Act preempts state action without matching that rigor, American kids could end up with weaker protections than their European peers.

That’s not a hypothetical. It’s a policy choice baked into the structure of the bill.

What Happens If Congress Ignores 44 State AGs

So what’s next? Congress could ignore the letter, pass the KIDS Act, and trigger a wave of lawsuits from states challenging federal preemption. That’s the most likely outcome if the bill moves forward as-is.

Alternatively, lawmakers could amend the bill to carve out space for state enforcement — a compromise that would preserve federal standards while allowing states to go further. That’s the pragmatic play, but it’s also the politically harder one, because it requires admitting the original bill was too industry-friendly.

The third option is that the bill stalls entirely. With 44 state AGs on record opposing it, the political cost of ramming it through just got a lot higher. Members of Congress who care about child safety — and who face reelection campaigns — might think twice about backing a bill their state’s top law enforcement officer publicly opposes.

Watch how the tech industry responds. If major AI and social media companies start lobbying harder for the KIDS Act, that’ll tell you everything you need to know about who the bill really protects. Platforms have a long history of pushing for federal preemption when state laws get too aggressive. This letter suggests states are done playing along.

Also watch whether any of the 44 AGs break ranks. A coalition this large is inherently fragile. If even a handful of states flip — especially large ones like Texas or Florida — the political calculus shifts. But as of now, the wall is holding.

Finally, keep an eye on the courts. If the KIDS Act passes and states sue, we’re looking at years of litigation over the scope of federal preemption. That’s a mess for everyone — except the lawyers.