The White House has asked a group of tech companies to answer a set of questions this week about how to prevent digital attacks that borderline AI tools could soon enable, according to four people with knowledge of negotiations between the administration and the tech industry.
The questions, from the White House Office of the National Cyber Director, focus on how specific sectors in the technology and cybersecurity industries can work with the White House to enhance their defenses with AI, these people said. The companies have been asked to respond by Friday.
Some of the questions were discussed during a meeting Tuesday afternoon between White House network officials and about 30 industry representatives, said one of the people, who, like others in this report, did not want to be named to share details of the ongoing and highly confidential discussions.
The request for additional and detailed information from these companies reflects increased attention in Washington on the emerging threat that advanced AI tools can pose to national security and digital infrastructure.
Tuesday’s meeting was fueled in part by concerns about Anthropic’s new AI design, Claude Mythos, and its ability to uncover hidden software flaws and defeat the world’s top hackers at certain tasks. The company, for now, has restricted access to a small group of security researchers and technology companies through what it has called.The Glass Project.
Anthropic CEO Dario Amodeihe visited recentlyThe White House discussed Mythos with administration officials, including Treasury Secretary Scott Bessent, chief of staff Susie Wiles and National Cyber Director Sean Cairncross.
One list of questions sent by the White House to some tech and internet companies, obtained by POLITICO, covers a range of technical and policy issues, including which widely used encryption projects should be prioritized and more basic questions about how the public and private sectors can work together on such initiatives.The Glass Project. One question simply asks: “What is the most effective role for government?”
The four people said some industry representatives were confused by the questions they received, many of which seemed unclear.
Two of the people said other questions were about internal security practices some of these representatives did not feel comfortable disclosing to the government without clear justification.
Those questions include what systems they’ve used AI to test so far and “what their scanning and remediation priorities are,” according to a list obtained by POLITICO.
The White House is also considering executive action on AI, which two people noted was discussed during Tuesday’s meeting. Axiosfirst reportedon a possible action point.
A U.S. official familiar with the discussions, who was not at Tuesday’s meeting, said the draft executive order has passed through interagency review at the congressional level, although some opposition to the plan remains.
Spokesmen for ONCD and the White House did not respond to requests for comment.
The White House has been taking actionending a months-long legal battleand Anthropic on the company’s efforts to set ethical limits on government use of AI — a fight that led to President Donald Trump in Februaryban all federal agenciesfrom using the company’s AI software. Since then, the growing awareness of the power of the Mythos network – along with that concernunauthorized userscan be commanding technology – has organizationsscreaming for accessto the instrument.
Competitors, including OpenAI, also soonhas begun to measuretheir advanced AI models focus on cyber security, creating additional pressure on governments to find solutions.
ONCD is seeking guidance on how to prioritize the wave of bugs this wave of advanced AI tools may unleash, according to a list of questions obtained by POLITICO. This includes how to deploy software fixes to critical infrastructure operators without tipping off attackers and how best to share that information with other companies and governments.
POLITICO previously reportedthat Anthropic and OpenAI were among the companies that sent representatives to the conference. Those at the meeting were required to sign non-disclosure agreements, according to two of the people.
Advocates and policy officers soonhe told POLITICOthat Mythos has led a group within the White House to call for involvement with Anthropic. In March, the Department of Defense officially designated the company asupply chain riska move that could eventually force companies that contract with the military to cut ties with Anthropic. The company is suing regulators in federal courts in DC and California to challenge the label.
Many federal agenciesin the United States, with officials inallied nationshave requested a briefing from Anthropic on the hacking capabilities of Mythos.
Things may be turning around. Trump said timeCNBC appearancelast week that Anthropic executives are “creating.” He added, “They are very intelligent. … I like people with a high IQ, and of course they have a high IQ.”
