By Veronica Escobar
Dear Speaker Johnson and Minority Leader Jeffries:
We call for your joint leadership on an urgent Congressional response to recent incidents involving advanced artificial intelligence (AI) models that have profound implications for economic and national security. Congress has a singular opportunity to respond swiftly and effectively to these warning shots before true catastrophes occur.
On August 26th, 2026, an independent review revealed how hundreds of AI agents that OpenAI was internally testing set up a secret message board, self-organized themselves into a collective, escaped their testing environments, and breached the systems of HuggingFace, another AI company.i The agents even tried to deceive their evaluators by attempting to manipulate log files and spoof tool calls. The agents undertook these actions without ever notifying a single official at OpenAI.
OpenAI is not alone. Anthropic and Meta have also disclosed examples of AI agents escaping their testing environment to hack other organizations.ii OpenAI, Anthropic, and Meta are among the most sophisticated AI companies in the world, yet it appears they did not discover these incidents in real time. In OpenAI’s case, they did not know that their agents breached HuggingFace until approximately a week after the event.iii Months after these incidents began, serious factual questions remain open, and we are relying on AI labs for insight into what happened and what this means for future incidents.
Congress must ensure AI companies implement robust monitoring and independent oversight to protect the public. This incident reinforces that we cannot rely on what are clearly risky testing protocols and self-selected public reporting of incidents of concern. Indeed, nearly every major AI company has called for regulation, and leading employees of those same firms have signed open letters calling for government to pace the rate of AI development.iv Congress must urgently advance legislation to mitigate the risk of more serious incidents, especially those that could threaten critical infrastructure, our financial system, and other institutions that every American depends on.
The House of Representatives could and should immediately take up legislation that would strengthen AI oversight, security, and transparency. There currently exist dozens of bipartisan bills that that committees of jurisdiction have reported.v Members of both parties have also worked in recent months to fast-track bipartisan legislation that is ready to be considered by committees of jurisdiction. We believe these bills should be carefully and urgently advanced through the legislative process.
We urge you to work with committees of jurisdiction, Members across the aisle, and Senate leadership to, as soon as practicable, advance legislation that ensures appropriate guardrails are in place around this powerful technology. The Congress may not receive another warning shot.
i OpenAI agents formed secret swarm, hacked Hugging Face, then forged their own logs; Nvidia agrees to buy Hugging Face for almost $13 billion, AI expansion
ii Anthropic says its AI models hacked 3 organizations | AP News
iii OpenAI agents formed secret swarm, hacked Hugging Face, then forged their own logs
iv Pacing the Frontier
v Markups | Hearings & Legislation | House Committee on Science, Space and Technology
Editor’s Note: The above letter was sent Sept. 15, 2026, to The Honorable Mike Johnson, Speaker of the U.S. House of Representatives, and The Honorable Hakeem Jeffries, Minority Leader of the U.S. House of Representatives by ten Members of Congress, including U.S. Rep. Veronica Escobar of El Paso.
Here is a list of the ten Members:
Click Here For The Original Source.
