A bipartisan group of representatives have asked House leadership to quickly take up and advance legislation that would establish guardrails for artificial intelligence after reports about models escaping their testing environments have fueled anxieties about the technology.
In a letter to House Speaker Mike Johnson and Minority Leader Hakeem Jeffries, a group of 10 representatives warned that recent incident disclosures from companies like OpenAI, Anthropic and Meta have “profound implications for economic and national security.”
That letter called for the House to “immediately take up legislation that would strengthen AI oversight, security, and transparency,” pointing to dozens of bipartisan bills that have been reported on those topics. The lawmakers specifically called for a focus on mitigating risk to areas such as critical infrastructure, the financial system and other institutions that Americans depend on.
“Congress has a singular opportunity to respond swiftly and effectively to these warning shots before true catastrophes occur,” they wrote.
Multiple AI policy heavy-hitters in the House were among those who signed the letter dated Tuesday and released publicly Wednesday. It was led by Rep. Don Beyer, D-Va. and signed by Reps. Jay Obernolte, R-Calif., Lori Trahan, D-Mass., Ted Lieu, D-Calif., Scott Franklin, R-Fla., Sara Jacobs, D-Calif., Gabe Amo, D-R.I., Valerie Foushee, D-N.C., Brian Fitzpatrick, R-Pa., and Veronica Escobar, D-Texas.
The correspondence follows news last week that Sen. Josh Hawley, R-Mo., is leading an investigation into the incident in which OpenAI models breached Hugging Face during a test, as well as allegations of the risks posed by AI. Those allegations have mounted in recent days after a former Anthropic researcher claimed that AI developers are “gambling with our lives” in the creation of the technology.
Notably, bipartisan action in the Senate and calls to act in the House come as President Donald Trump is denying issues with the technology. In a post to his social media site Monday, he claimed the fears about AI were a “hoax.”
In addition to calling for swift action, the letter also took aim at the failure of AI companies’ own testing protocols. For example, OpenAI disclosed that it was unaware that its model had hacked Hugging Face for roughly a week, the letter said.
“This incident reinforces that we cannot rely on what are clearly risky testing protocols and self-selected public reporting of incidents of concern,” the lawmakers wrote.
While the letter underscored urgency in the matter, the chamber is about to enter into a recess that will run through the midterm elections. That recess is starting a day early after Johnson canceled votes Thursday. “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 letter said. “The Congress may not receive another warning shot.”