# Trump Signs Executive Order on AI Model Reviews > Source: > Published: 2026-06-02 20:42:09.035556+00:00 # Trump Signs Executive Order on AI Model Reviews President Donald Trump signed an executive order titled "Promoting Advanced Artificial Intelligence Innovation and Security" on June 2, 2026, the White House published. The order asks AI developers to voluntarily submit their most capable models for government cybersecurity review for up to **30 days** before public release, according to the White House text and reporting by Reuters and NPR. The directive directs Treasury, Defense, Commerce and Homeland Security to coordinate testing and to create an AI "cybersecurity clearinghouse," per the White House and Reuters. Reporting by The New York Times and Reuters frames the move as a departure from the administration's prior hands-off approach; earlier drafts reportedly contained a **90-day** review window that was cut to **30 days**, according to multiple outlets. ### What happened President Donald Trump signed an executive order titled "Promoting Advanced Artificial Intelligence Innovation and Security" on June 2, 2026, according to the White House text. The order asks AI developers to voluntarily submit their most capable models for government cybersecurity review for up to **30 days** before public release, as described in the White House text and reported by Reuters and NPR. The order assigns coordination roles to the departments of Treasury, Defense, Commerce and Homeland Security and directs the Treasury secretary to help form an AI "cybersecurity clearinghouse," per the White House document and reportage in Newser and Reuters. Multiple outlets, including The New York Times and Reuters, report that an earlier version of the directive had proposed a **90-day** review period that was shortened to **30 days** in the final order. ### Technical details Editorial analysis - technical context: Public reporting states the order tasks federal agencies with developing benchmarks to assess AI models' cyber capabilities and with negotiating voluntary testing agreements with developers (Reuters, NPR, White House text). The White House text includes operational deadlines for agency action within **30 days** in several sections, and it explicitly states that "nothing in this section shall be construed to authorize the creation of a mandatory governmental licensing, preclearance, or permitting requirement for the development, publication, release, or distribution of new AI models," language echoed in Newser's coverage. ### Context and significance Reporting by The New York Times and Reuters frames the order as a shift away from the administration's previously hands-off stance toward AI. Multiple outlets link the move to growing national-security and cybersecurity concerns about more capable models; NPR and Reuters note recent industry moves and announcements, including limited releases by some firms, as part of the background that influenced debate. The order is the administration's most visible AI policy action to date, according to The New York Times. Editorial analysis - practitioner implications: For AI practitioners and security teams, the order formalizes government interest in model-level testing and vulnerability assessment without creating an explicit licensing regime, per the White House text. Observers following comparable voluntary frameworks note that formalizing benchmark development and an information-sharing clearinghouse tends to increase coordination between vendors and government but can also introduce process overhead around release timelines. ### What to watch - •Whether major developers sign voluntary testing agreements and how they define "most capable models," as reported discussions between agencies and firms were referenced by Reuters. - •How the Treasury-led clearinghouse will operate in practice: what data it will require, retention and sharing policies, and its interfaces with CISA and Defense, following mandates in the White House text. - •The benchmarks agencies adopt for "cyber capabilities" assessments and whether those benchmarks focus on exploitability, model hallucination leading to security risk, or adversarial misuse, a key distinction for practitioners designing evaluations. Observed patterns in similar transitions: When governments propose voluntary review windows and information-sharing mechanisms, industry participants often negotiate scope, timelines, and confidentiality protections; those discussions typically shape the operational details practitioners must implement. Reporters note that the shorter **30-day** window reflected compromise during White House deliberations, according to The New York Times and Reuters. ### Bottom line Reporting indicates the executive order creates a voluntary, centralized pathway for government cybersecurity review of advanced models and establishes agency responsibilities and benchmarks, while explicitly avoiding a statutory licensing regime, per the White House text and multiple news reports. The policy converts mounting national-security concern about advanced AI into concrete interagency tasks and a Treasury-led clearinghouse, with the practical impacts set to depend on follow-on agreements and implementation details. ## Scoring Rationale The order is a notable federal policy step that creates a cross-agency mechanism for voluntary model review and a clearinghouse, directly affecting security practices and vendor-government interactions. The impact depends on implementation and industry uptake, so the story rates as a significant but not paradigm-shifting development for practitioners. Practice interview problems based on real data 1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with. [Try 250 free problems](/problems)