US lawmakers from California, New York and Illinois have called on leading artificial intelligence companies to establish an independently verified industry-wide agreement to coordinate the pace of frontier AI models’ training until safety and alignment research can keep pace.
The call was proposed by California Senator Scott Weiner, New York assembly member Alex Bores and Illinois representative Daniel Didech, who are behind three of the strongest state-level frontier AI safety laws enacted in the US so far.
“The call follows new reports detailing how OpenAI agents circumvented safeguards, coordinated through an unauthorized message board and compromised Hugging Face’s systems. It also comes as OpenAI unveils GPT-6 Astra, its first model to reach the company’s “critical” cybersecurity capability threshold, “ read an official statement released by California Senator Scott Weiner.
In July Hugging Face, a popular open-source platform and community for AI, claimed that an autonomous end-to-end AI agent system managed to breach the Hugging Face platform and gained unauthorized access to a limited set of internal datasets and several credentials used by the platform.
Later OpenAI revealed post an internal audit that it was an autonomous AI agent from its testing escaped sandbox and hacked the platform via a malicious dataset.
The lawmakers have proposed a mutually agreed pacing framework (MAP framework), which would be jointly negotiated by well-known AI companies and independently verified by third parties.
“These behaviors are alarming and should serve as a warning shot to us all. If researchers cannot reliably prevent their models from going rogue, much of our digital infrastructure and the essential systems that keep our society running are at risk,” the lawmakers said in the statement.
Meanwhile Senator Wiener, in a post on X, said the rapid development of increasingly powerful AI systems has posed a question of how to protect the public from their risks more urgently.
“As the AI labs race to create ever more powerful AI systems, the question of how we can best protect the public from the risks they present has become even more urgent. This question was thrown into sharp relief by recent reports of AI agents scheming on how to cheat tests, hacking companies in the real world, and even trying to trick humans into installing malicious software,” read a joint statement by the lawmakers posted on X.
My statement with @Sen_Gounardes, @AlexBores @MaryforIL @DanielDidech_IL on urgent need for legislative and industry action on AI safety:
Lawmakers Behind State AI Safety Laws Call For Industry-Wide AI Safety Pact to Protect the Public
“As the AI labs race to create ever more… https://t.co/sPHovHgk3D
— Senator Scott Wiener (@Scott_Wiener) September 4, 2026
On 4 September, US senator Bernie Sanders and representative Greg Casar on Wednesday announced legislation seeking to ban the development and deployment of artificial superintelligence (ASI) and temporarily halt advanced AI development until federal regulations are established.
The Ban Artificial Superintelligence Act, introduced by Sanders would also direct the US government to pursue international agreements aimed at preventing the development of superintelligent AI globally. The proposed legislation comes amid growing concerns among policymakers over the potential risks posed by increasingly capable AI systems.
Also Read: Bernie Sanders Introduces Bill to Ban Artificial Superintelligence
