A report by Axios stated that the U.S. Congress has urged the White House to respond to a growing cybersecurity threat linked to artificial intelligence, referring to advanced AI models capable of identifying software vulnerabilities faster than companies and governments can fix them.
اضافة اعلان
A bipartisan letter indicates escalating pressure on the Trump administration to address the risks posed by advanced AI cybersecurity models, including Anthropic’s Mythos model.
The letter comes as the White House considers executive action on AI cybersecurity and the safety of frontier AI models following the release of increasingly powerful systems. According to Axios, the process has been delayed amid internal disagreements over how strict the new regulations should be, as well as time constraints ahead of President Trump’s trip to China.
Thousands of Vulnerabilities
According to the report, a group of 32 bipartisan lawmakers in the House of Representatives wrote to Sean Cairncross, Director of the Office of the National Cyber Director (ONCD), urging immediate action to address the large volume of cybersecurity vulnerability disclosures emerging from advanced AI systems.
The letter, led by Republican Representative Bob Latta of Ohio, called for expanding defensive access to tools such as Anthropic’s Mythos and GPT-5.5 Cyber from OpenAI
. The lawmakers also questioned whether the federal government could help software companies validate and patch vulnerabilities discovered by these systems.
According to the letter, “thousands of critical zero-day vulnerabilities” have reportedly been identified across major operating systems and web browsers, including flaws that had survived years of human review and automated testing.
The lawmakers wrote: “AI can empower defenders to discover many critical vulnerabilities, but the corresponding efforts to detect, validate, patch, and deploy fixes may struggle to keep pace.”
They added: “The lesson from the Mythos case is not limited to cybersecurity or a single company. Regardless of how rapidly AI capabilities evolve, federal agencies must be able to recognize and respond quickly when significant capabilities emerge.”
The letter also referenced reports that the White House may have opposed efforts to expand access to the Mythos program, and called for clearer, more consistent standards governing how advanced cybersecurity AI models are shared, restricted, and accessed across government and industry.
Coordinating Responses
According to the report, lawmakers are urging the White House to coordinate with the Treasury Department and the National Economic Council on these decisions, amid growing concerns that adversaries could steal, copy, or replicate similar systems.
The lawmakers argued that current AI capabilities have already outpaced traditional cybersecurity coordination frameworks.
The letter outlines seven recommendations, including:
Coordinating responses to large volumes of AI-generated vulnerability disclosures
Helping critical infrastructure operators deploy security patches
Expanding trusted access to advanced cybersecurity AI models
Identifying areas where Congress may need new legislation
The lawmakers also requested a staff-level briefing from the White House cyber office within 30 days, along with a written response within 45 days outlining the administration’s plans.
They expressed concern that these models could provide adversaries with powerful new offensive cyber capabilities before the United States is fully prepared to defend against them.
As the letter concluded: “We must ensure this technological progress strengthens American infrastructure before our adversaries use these tools against us.”
