President Donald Trump formally signed an executive order focused on the oversight of artificial intelligence (AI) on Tuesday, marking a significant step for federal government engagement with advanced technological development. This action came less than two weeks after the President had notably postponed a planned White House ceremony, driven by his concerns that an earlier version of such a policy might inadvertently stifle America’s crucial technological lead on the global stage.

The newly signed executive order establishes a structured framework designed to enable the federal government to meticulously vet the national security implications and risks associated with the most cutting-edge AI systems. This review process, which precedes the public release of these advanced systems, is stipulated to last for a maximum of 30 days. A key characteristic of this policy is its voluntary nature, clearly stating that participation by AI developers in this vetting process is entirely optional. The order explicitly articulates its foundational principle: “Advanced AI capabilities make our Nation stronger, but also introduce new national security considerations that require coordinated action across executive departments and agencies.”

Sources indicated that it was not immediately clear how extensively the final order differed from the version Trump had declined to sign on May 21. However, the specified 30-day window for government review of an AI system was notably shorter than what some industry stakeholders had anticipated. This deliberate choice of a concise review period was likely a strategic decision to prevent the policy from becoming an undue burden on a rapidly evolving and intensely competitive industry. President Trump had previously canceled an Oval Office event with tech industry executives, articulating his apprehension: “We’re leading China, we’re leading everybody, and I don’t want to do anything that’s going to get in the way of that lead,” he had told reporters at the time, underscoring his priority of preserving America’s technological preeminence.

The directive is predicated on a voluntary collaborative model involving prominent U.S.-based technology companies. These firms, often referred to as "frontier labs" due to their pioneering work in developing the most advanced AI systems, include industry giants such as Anthropic, OpenAI, and Google. Several executives from these companies had originally intended to be present at the May 21 signing ceremony, which was ultimately conducted without any public fanfare. The White House, in a social media communication on Tuesday, further clarified the scope of the order, stating that it “creates a process for frontier labs to voluntarily share cutting-edge cyber models in order to secure critical infrastructure and strengthen the government’s own cyber defenses. We are NOT conducting oversight of all new models, as that level of government overreach would have chilling effects on free speech and innovation.” This clarification aimed to assuage concerns about potential excessive governmental control over AI development.

Expert analysis of the order presented a mixed but generally positive reception. Juan Londoño, a policy analyst at the libertarian-leaning Cato Institute, characterized the order as imperfect but nonetheless “a step in the right direction to prepare the nation for the release of advanced AI systems.” While commending the White House’s emphasis on the voluntary nature of the process, Londoño expressed significant concerns regarding the inherent vagueness in how the government, under the leadership of the director of the National Security Agency (NSA), would definitively determine which specific AI models would qualify for scrutiny. He also highlighted the lack of clarity on the criteria for deciding which "trusted partners" would gain early access to these systems. In an interview, Londoño warned that granting such extensive discretion to the NSA director could establish a “dangerous precedent,” potentially enabling the government to “weaponize” the policy against technology companies with which it might have existing disputes, citing Anthropic as a specific example.

The impetus for this new AI cybersecurity directive notably followed Anthropic’s April announcement of its highly advanced AI model, "Claude Mythos." This revelation occurred amidst the company’s ongoing legal entanglement with the Trump administration, stemming from a contract dispute with the Pentagon. The potential risks posed by Mythos were quickly recognized; Treasury Secretary Scott Bessent and the outgoing Federal Reserve Chair Jerome Powell soon thereafter convened an urgent meeting with leading Wall Street CEOs. During this meeting, they issued stark warnings about Mythos’s apparent capability to identify significant cybersecurity vulnerabilities within global software infrastructures. Anthropic had initially restricted access to Mythos to an exclusive group of "trusted partners," primarily comprising major tech companies and financial institutions. However, the company announced on Tuesday that it had since expanded this group to include an additional 150 organizations, broadening its reach while maintaining controlled access.

The executive order garnered a largely favorable response from the very tech giants it seeks to engage. Anthropic itself lauded Trump’s new policy as “an important step in strengthening America’s leadership in AI” and affirmed its readiness to collaborate with the White House on its effective implementation. Similarly, OpenAI, the creator of the rival ChatGPT, and Google both echoed this sentiment, describing the policy as an important advancement. Chris Lehane, OpenAI’s chief global affairs officer, emphasized the company’s perspective: “As AI capabilities continue to advance, we believe effective safety frameworks should continue to be developed through democratic institutions, informed by technical expertise and broad stakeholder input, to promote accountability and public trust.”

On the political front, Democratic Senator Mark Warner, who serves as vice chairman of the Senate Intelligence Committee, also welcomed Trump’s policy. However, he concurrently issued a critical assessment of the administration, noting that it had “belatedly discovered the need to redo something it hastily dismantled in its first year.” This remark alluded to Trump’s previous actions, specifically his repeal of many of former President Joe Biden’s established guardrails for AI shortly after returning to the White House the previous year. AP Technology Writer Barbara Ortutay also contributed to the reporting on this significant development.