Full Breakdown
U.S. Commerce Department Removes AI Security Testing Details
5/11/2026, 11:21:56 PM
Removal of AI Security Testing Announcement
On May 11, 2026, the U.S. Commerce Department deleted the webpage that had outlined its agreement with Google, xAI, and Microsoft to provide early-access AI models for government security testing. The former link now shows a “Sorry, we cannot find that page” notice and redirects to the Center for AI Standards and Innovation (CASI), the agency overseeing the assessments.
Background and Timeline
On May 5, the Commerce Department announced that the three companies would hand over new AI models before public deployment so that federal scientists could examine them for vulnerabilities that might enable cyberattacks or military misuse. The announcement listed Google, Elon Musk’s xAI, and Microsoft as participants and cited growing concerns about powerful systems such as Anthropic’s Mythos. Six days later, the detailed announcement page was removed, and the link now leads to CASI’s site.
Implications for National Security
Early access to advanced models allows U.S. officials to identify security flaws that could threaten critical infrastructure or be weaponized, thereby safeguarding national security. The disappearance of the public notice reduces transparency about the program’s scope and timeline, raising questions about how the government will continue to monitor emerging AI risks.
Official Statements & Responses
Spokespeople for both the Commerce Department and the Trump White House immediately declined to comment when approached by Reuters. No alternative official communication has been released regarding the page removal.
Criticism & Opposition
Government insiders have warned that powerful AI systems pose national-security risks. The reference to Anthropic’s Mythos underscores the breadth of models under review and the urgency of establishing robust security standards.
Conflicting Reports & Gaps
The specific reason for deleting the webpage remains unclear. While the May 5 announcement described the testing framework, the current absence of the page leaves a gap in publicly available details about the program’s operational parameters.
Verbatim Quotes
- “Sorry, we cannot find that page.” — Commerce Department website message
- “It was not immediately clear why the website was deleted.” — Reuters report
- “government over the national security risks posed by powerful AI systems, including Anthropic's Mythos.” — Reuters report
- “Spokespeople for the Commerce Department and Trump White House did not immediately respond to a request for comment from Reuters.” — Reuters report
What’s Next
The Commerce Department has not indicated whether the testing agreement will proceed unchanged or when public information about the program might be restored. Ongoing monitoring by journalists and policymakers will likely focus on any future disclosures from CASI or the participating companies.
