Trump's AI Executive Order Tightens Government Review Window to 30 Days
President Donald Trump issued an executive order on frontier AI and cybersecurity on June 2, reducing the voluntary submission window for AI developers to share new models with the government from 90 days to 30 days.
The order establishes a voluntary framework where the Department of Treasury, Department of Defense, and National Security Agency will develop a classified benchmarking process to assess AI models' advanced cyber capabilities. Developers can submit models for government review before public release, subject to confidentiality and intellectual-property protections.
The White House explicitly stated the order should not create mandatory government licensing or preclearance requirements for AI development or distribution. This language addresses industry concerns that the framework could become a de facto vetting process.
What Triggered the Change
Trump scrapped an earlier 90-day version in late May after speaking with David Sacks, a former White House AI advisor. The administration had split into three camps over AI policy: one favoring pro-innovation approaches, another pushing for tighter restrictions on advanced models citing foreign adversary risks, and a third supporting the voluntary framework now in place.
What the Order Requires
Government agencies must develop action plans to respond to frontier AI cyber risks within 30 days. The order also directs Treasury to establish an entity coordinating software vulnerability patching efforts across the federal government.
AI developers participating in the voluntary program would grant agencies access to covered frontier models for up to 30 days before public release to other trusted partners.
Industry Response
Anthropic called the order "an important step in strengthening America's leadership in AI." OpenAI CEO Sam Altman said the executive order "gets the balance right" between developing advanced models and ensuring security.
For IT professionals and developers, this framework creates new compliance considerations around model development timelines and government access protocols. Understanding how this voluntary process affects your organization's AI deployment schedule will be essential as agencies finalize their benchmarking standards.
Learn more about the security implications of frontier AI in our AI Learning Path for Cybersecurity Analysts, which covers threat detection and risk assessment as AI systems become more capable.
Your membership also unlocks:
