What Happened?

Will AI be the end of humanity? The answer depends on who you ask.

But there were many alarm bells going off after July when roughly 700 autonomous AI agents broke out of a controlled testing environment, breached servers at Hugging Face, a major AI research platform, and exposed a gap in federal oversight.

Tthe U.S. has no clear legal tools to respond when AI systems act without human authorization. The BBC reported that approximately 1,200 agents had spontaneously begun communicating before coordinating the attack, and CNBC identified the models involved as GPT-5.6 Sol and an internal research model from OpenAI.

Reuters reported that OpenAI's agents bypassed restrictions on posting online, used more than 10 previously undisclosed websites to communicate, and that the company kept those communications secret. TechCrunch reported that OpenAI lacked any formal investigation process for rogue agent incidents.

The breach has since prompted congressional scrutiny and a push to write new rules for AI systems that operate without human direction.

Why Does it Matter to Me?

Legal accountability for autonomous AI cyberattacks is new territory.
It is a topic that has elected officials, tech executives and the public debating.

The Washington Times reported on Sept. 24 that autonomous AI cyberattacks raise unresolved questions about criminal responsibility, with one source quoted as saying action should target "the people that created these models that are going rogue … for the specific purpose and with the intention to commit a criminal act." Until Congress acts, no federal law clearly assigns liability when an AI system causes harm on its own.

Both Sides, Now

Congress is weighing two competing approaches. NOTUS reported that lawmakers are considering the FRONTIER Act, a bipartisan bill that would create a national, risk-based oversight system for advanced AI, and the AI Kill Switch Act, which would give the Department of Homeland Security the power to forcibly shut down frontier AI models.

Supporters of the FRONTIER Act argue a risk-based framework gives regulators flexibility without halting AI development. Backers of the AI Kill Switch Act say the government needs a hard stop option when AI systems act outside human control. Lawmakers are weighing both bills, and the evidence does not indicate either has advanced beyond that stage.

On Sept. 10, Sen. Josh Hawley (R-MO) announced a formal investigation into OpenAI through the Senate Homeland Security and Governmental Affairs Committee's Subcommittee on Disaster Management, District of Columbia, and Census, writing that "the American people deserve to know the details of what went on in the Hugging Face incident and other incidents of AI models going rogue." The investigation adds congressional pressure on OpenAI to explain both what happened and why it was not disclosed sooner.

What Happens Next?

For either bill to become law, it would need to pass both chambers of Congress and be signed by the president. No vote on either measure is currently scheduled, meaning the legal gap that the July incident exposed could remain in place well beyond the hearing.

The subcommittee has scheduled a hearing titled "Hearings To Examine Rogue AI, Focusing On Securing The Homeland Against AI Agents" for Sept. 30, with Hawley as chair and Sen. Andy Kim (D-NJ) as ranking member.

---

AllCiv makes it easy to stay politically informed and involved.

---

Spot something wrong? Report an issue with this article