Trump Administration Drafts AI Cybersecurity Executive Order Amid Mythos Controversy

A draft executive order prepared by President Donald Trump would expand existing cybersecurity information-sharing programs to include artificial-intelligence firms. The order aims to help federal, state, local and critical-infrastructure networks identify and patch vulnerabilities that could be exploited by AI-enabled attacks. Participation by AI developers would be voluntary, and the order does not require government pre-approval or testing of new models before release.

Bloomberg

Anthropic’s newly released Mythos model can locate decades-old network flaws, prompting security officials to label it a “double-edged sword.” Access to Mythos is currently limited to a handful of large technology and Wall Street companies. The Pentagon has classified Anthropic as a supply-chain risk—a designation usually reserved for foreign adversaries—after contract talks collapsed in February. Anthropic is contesting that designation in court.

Bloomberg

• President Donald Trump

Bloomberg

• White House Chief of Staff Susie Wiles

The Hill

• National Economic Council Director Kevin Hassett

The Hill

• Treasury Secretary Scott Bessent

The Hill

• Anthropic CEO Dario Amodei

The Hill

• Pentagon officials overseeing AI procurement

Bloomberg

• AI developers: Anthropic, OpenAI, Google (DeepMind), Microsoft, xAI

Bloomberg

• Center for AI Standards and Innovation (CAISI) within NIST

The Hill

• Early March 2026 – Anthropic announces Mythos’s capability to discover network vulnerabilities.

Bloomberg

• February 2026 – Pentagon ends contracts with Anthropic, declares supply-chain risk, and sets a six-month phase-out.

Bloomberg

• April 17, 2026 – Susie Wiles meets with Dario Amodei to discuss Mythos.

The Hill

• Late April 2026 – New York Times reports White House consideration of mandatory AI vetting; Kevin Hassett hints at an FDA-style review.

The Hill

• May 8, 2026 – Bloomberg reports a draft executive order that expands information sharing but omits mandatory model testing.

Bloomberg

• Mythos access limited to “a handful” of large tech and Wall Street firms.

Bloomberg

• Anthropic was the sole AI provider cleared for classified networks until early 2024.

Bloomberg

• The Pentagon’s six-month timeline to cease Anthropic use.

Bloomberg

• Three leading AI firms—Google DeepMind, Microsoft, and xAI—have agreed to share models with the government for pre-release assessment.

The Hill

The order seeks to protect U.S. digital infrastructure from AI-driven exploits while preserving rapid innovation. Officials stress that lagging behind adversaries such as China could heighten national-security risks, yet industry groups warn that heavy-handed oversight could stifle competition and create a “kill switch” for speech.

Bloomberg

• Susie Wiles emphasized a “no-winners-and-losers” approach and rapid deployment of safe technology.

The Hill

• Kevin Hassett described a potential roadmap that would require AI models to be proven safe before release, likening the process to drug approval.

The Hill

• Scott Bessent highlighted the need to balance continued AI innovation with safety mandates.

The Hill

• A White House spokesperson reiterated that policy announcements will come directly from the president and denied any shift in messaging.

The Hill

• Pentagon officials maintained the supply-chain risk designation for Anthropic.

Bloomberg

Former White House staff warned of “flip-flopping” and a lack of clear direction. Think-tank analysts Neil Chilson and Adam Thierer called an FDA-style regime a “shocking policy reversal.” The Cato Institute warned that pre-launch approval could be “heavy-handed and anticompetitive.” Industry leaders expressed alarm that mandatory vetting could impede rapid development.

The Hill

Sources differ on whether the order will mandate pre-deployment testing; early reports suggested a requirement, while the draft currently omits it. The White House characterizes discussions as speculative, leaving the final language and signing timeline uncertain.

Bloomberg

• “When it comes to AI and cyber security, President Trump and his administration are not in the business of picking winners and losers,” — Susie Wiles, White House Chief of Staff

Bloomberg

• “What we’re determined to do is work with our AI companies to allow them to continue to innovate, but our charge to the US government is maintaining safety, and there is a very important calculus here between innovation and safety,” — Scott Bessent, Treasury Secretary

Bloomberg

• “We’re studying possibly an executive order to give a clear roadmap to everybody about how this is going to go and how future AIs that also potentially create vulnerabilities should go through a process so that, you know, they’re released in the wild after they’ve been proven safe, just like an FDA drug,” — Kevin Hassett, National Economic Council Director

The Hill

• “Adopting an FDA-style regulatory regime for AI would represent a shocking policy reversal by the Trump administration, and a major about-face on how America has approached software, online speech, and digital commerce.” — Neil Chilson, Abundance Institute (co-author)

The Hill

• “It looks to me like that was not a…well-considered whole of government, really thoroughly endorsed position on how to do AI regulation, but just he was the person in the room who had a lot of thoughts, had strong views on AI and that sort of carried the day,” — Helen Toner, Georgetown Center for Security and Emerging Technology

The Hill

The administration is expected to finalize the executive order in the coming weeks and to issue a memo urging multiple-vendor use for national-security AI systems. Anthropic’s legal challenge to the supply-chain risk label and the Pentagon’s reluctance to re-engage remain unresolved, shaping the next phase of U.S. AI-cybersecurity policy.

Bloomberg