TL;DR: The Senate Judiciary Committee voted 22-0 on April 30, 2026 to advance the GUARD Act (S. 3062), which requires every American to verify their identity (via government ID, facial scan, or financial records) before using AI chatbots. Minors would be banned entirely from AI companions. The bill carries penalties of up to $250,000 per violation. The EFF calls it "a surveillance mandate disguised as child safety." NetChoice warns it creates a "massive honeypot" of personal data. A companion House bill (H.R. 8623) was introduced the same day. Both chambers have bipartisan support.
What the Senate Just Voted For
On April 30, 2026, the Senate Judiciary Committee advanced the GUARD Act (Guidelines for User Age-verification and Responsible Dialogue Act) without a single dissenting vote. Twenty-two senators. Zero opposition. Bipartisan unanimity on a bill that would require identity verification for anyone who wants to talk to an AI chatbot.
The bill's sponsors frame it as child protection. Senator Josh Hawley (R-Missouri) and Senator Richard Blumenthal (D-Connecticut) want to keep kids away from AI companions that can simulate emotional relationships. That part sounds reasonable. The implementation is where things go sideways.
To verify that users are adults, the GUARD Act requires "reasonable age verification" tied to real-world identity. Simple checkboxes, birth date fields, and IP-based checks are explicitly banned. What's left? Three options:
- Government-issued photo ID. Upload a scan of your driver's license or passport.
- Facial scan. Submit biometric data to prove you match your ID.
- Financial records. Link a bank account or credit card to verify your age through financial history.
Every user gets verified. Not just kids. Not just suspicious accounts. Every single person who wants to ask an AI chatbot a question.
The Scope Is Wider Than You Think
Congress narrowed the bill after the original version could have applied to "nearly every AI-powered chatbot or search tool," including search engines. The revised definition targets systems that "engage in interactions involving emotional disclosures" from users or present a "persistent identity, persona or character."
That sounds narrow. It isn't.
Modern AI chatbots are designed to be conversational. They remember your name. They respond empathetically when you say you're frustrated. A customer service bot that says "I understand that must be stressful" could trigger the bill's definition. The EFF flagged this directly: as conversational AI becomes more emotionally responsive, the boundary between "AI companion" and "regular chatbot" dissolves.
Fight for the Future went further, calling the GUARD Act "a Trojan horse for universal online ID checks." Their concern: the bill could require verification from "everyone from internet service providers to anyone who runs a blog" with AI-powered features.
NetChoice spelled out the practical absurdity: "If a fast food restaurant or the website of an airline provides a chat feature," users would need to hand over identity documents just to ask about their order status.
The House companion bill, H.R. 8623, introduced April 30 by Representatives Blake Moore (R-Utah) and Valerie Foushee (D-North Carolina), mirrors the Senate version. Both chambers, both parties, same bill.
Building the World's Most Valuable Hack Target
Every age verification system collects sensitive data. The GUARD Act would require millions of Americans to upload government IDs, facial scans, or financial records to companies that build chatbots. That data has to go somewhere. It has to be stored. And it has to be secured by companies whose core competency is building AI, not protecting identity databases.
NetChoice's warning was blunt: age verification creates a "massive honeypot of personally identifiable information" for hackers. The bill requires constant re-verification, meaning "parents will have to consistently upload theirs and their children's personal data," multiplying the attack surface with every interaction.
This isn't theoretical. Age verification systems have already been breached. NetChoice cited the EU's age verification app as a cautionary tale: "After launch, it took security experts all of two minutes to hack the system." Two minutes.
Now imagine a database containing the government IDs and facial scans of every American who uses ChatGPT, Claude, or Google's AI tools. That's not just a privacy risk. That's a national security target.
We've covered Roblox deploying facial recognition for 100 million children through a third-party service called Persona. Those systems collect biometric data from minors. The GUARD Act would do the same thing, at a national scale, for adults and children alike.
Who Gets Locked Out
The bill assumes every American has a government ID, a bank account, or access to facial scanning technology. Millions don't.
The EFF pointed out that "millions of Americans do not have current government ID, accounts at major banks, or stable access to the kinds of digital identity systems the bill contemplates." That includes elderly Americans whose IDs have expired, homeless individuals, undocumented residents, and rural communities with limited banking access.
Over half of U.S. teenagers currently use AI chatbots for schoolwork. The GUARD Act would ban them entirely from AI companions and make it significantly harder for their parents to access the same tools. No parental consent option exists. No appeal mechanism for users flagged as underage.
The penalty structure makes this worse. Companies face fines up to $250,000 per violation, raised from $100,000 in the original bill, enforced by both the U.S. Attorney General and state attorneys general. With that kind of liability, the EFF predicts "smaller developers, in particular, may decide it is safer to block younger users entirely, disable conversational features, or avoid developing certain tools at all." The rational business response to vague rules and enormous fines is to over-comply. Block everyone who looks like a risk.
The people who need AI tools most (students researching homework, isolated individuals seeking conversation, people with disabilities using AI for daily assistance) are the first ones locked out.
Child Safety or Surveillance Infrastructure?
The GUARD Act follows a pattern we've tracked across multiple stories. Age verification laws in Utah, Louisiana, and other states have used child safety as the justification for building identity verification systems that apply to everyone. The logic is always the same: to keep children out, you have to check everyone's ID at the door.
The EFF titled its original analysis: "A Surveillance Mandate Disguised As Child Safety." Their argument: if Congress wanted to protect kids from harmful AI interactions, it would fund enforcement against bad actors and pass comprehensive privacy laws. Instead, the GUARD Act builds verification infrastructure first and asks questions later.
The bill also requires AI chatbots to disclose they aren't human at the start of every conversation and every 30 minutes thereafter. Chatbots are prohibited from claiming to be licensed professionals: doctors, lawyers, therapists, financial advisors. Those are sensible guardrails. They don't require identity verification.
That's the tell. The disclosure requirements and professional impersonation bans could exist as standalone legislation. Attaching them to a universal identity verification mandate reveals the priority. Congress isn't just worried about what chatbots say. It's interested in knowing who's listening.
The First Amendment Problem
NetChoice's opposition letter cited Brown v. Entertainment Merchants Association (2011), the Supreme Court case that struck down California's ban on selling violent video games to minors. The court held that the government can't restrict access to speech based on content (even for kids) without meeting strict scrutiny.
The GUARD Act doesn't just restrict access for minors. It conditions access for everyone on surrendering identity documents. That's a prior restraint on speech, requiring government-approved identification before you can engage with a communication tool.
NetChoice distinguished the GUARD Act from Free Speech Coalition v. Paxton (2025), where a court approved age verification only for illegal content like pornography. The GUARD Act covers legal speech: asking a chatbot to explain quantum physics or help draft an email. Requiring ID for that crosses a constitutional line that courts have drawn repeatedly.
Ashkhen Kazaryan of The Future for Free Speech called the bill exactly what it is: mandatory "government ID or equivalent age verification for any American who wishes to interact with an AI chatbot."
What You Can Do
Contact Your Senators
The GUARD Act is heading to a full Senate floor vote. The EFF has an action page where you can send a letter to your representatives opposing the bill. Twenty-two senators voted yes in committee. They need to hear from constituents before the floor vote.
Refuse Biometric Verification Where Possible
When services offer facial scanning as "convenient" verification, choose alternatives. Every face scan you submit creates another copy of your biometric data in another database. Unlike passwords, you can't change your face after a breach.
Support Privacy-First AI Tools
Use AI services that minimize data collection and don't require identity verification. Open-source models you can run locally, like Llama or Mistral, don't need to know who you are. The more people use privacy-respecting alternatives, the harder it is to normalize surveillance as a default.
Watch for State-Level Copycats
Age verification laws are spreading state by state. Utah, Louisiana, Virginia, and others have passed or proposed similar requirements. Your state legislature may introduce its own version. Track it through the EFF's legislative tracker or your state's bill database.
The Bottom Line
The GUARD Act sailed through committee with zero opposition. That's how popular "protect the children" legislation is, even when it requires every adult in America to hand over their government ID to talk to a chatbot.
The bill doesn't protect children from AI risks. It builds identity verification infrastructure that applies to every user, creates honeypot databases of sensitive documents, locks out millions of Americans who lack the required ID, and chills the development of AI tools through vague definitions and quarter-million-dollar penalties.
If this passes, the precedent is set. Today it's AI chatbots. Tomorrow it's search engines, messaging apps, or any digital service Congress decides needs an ID check. The infrastructure doesn't get dismantled after it's built. It gets expanded.
Congress knows how to protect kids online. Enforce existing laws. Fund oversight. Pass privacy legislation that limits what data companies collect in the first place. What Congress chose instead is an identity database. That tells you everything about the actual priority.
References
- EFF: Congress Narrowed the GUARD Act, But Serious Problems Remain (May 2026)
- EFF: The GUARD Act Isn't Targeting Dangerous AI. It's Blocking Everyday Internet Use (April 2026)
- EFF: A Surveillance Mandate Disguised As Child Safety (November 2025)
- NetChoice: Letter of Opposition to the GUARD Act (2026)
- IBTimes: US Senate Advances GUARD Act, AI Chatbot Age Verification (2026)
- Reason: How a bill banning AI companions for kids could usher in widespread ID checks online (May 2026)
- Global Policy Watch: Senate Judiciary Committee Advances GUARD Act (May 2026)
- Congress.gov: S.3062, GUARD Act Full Text
- Congress.gov: H.R.8623, GUARD Act House Companion Bill