Trump Expected to Sign New AI and Cybersecurity Executive Order
US President Donald Trump is anticipated to sign an executive order concerning AI and cybersecurity as early as Thursday, according to sources, amid rising pressure from segments of his political base to enhance regulation of emerging AI models, including Anthropic’s Mythos. The White House was coordinating efforts to bring AI company CEOs to a signing ceremony with President Donald Trump, according to another source. The order would establish a voluntary framework for AI developers to interact with the US government regarding the public release of covered models, according to sources. Under the framework, developers would be required to submit their models to the government 90 days prior to public release, and also grant pre-public access to essential infrastructure providers, including banks, according to one source. Such an approach may signify a compromise position among Trump supporters. MAGA activists, including former Trump adviser Steve Bannon and right-wing political organiser Amy Kremer, have been advocating for the White House to mandate that AI developers submit their most advanced models for government security evaluations. Conversely, proponents from the tech sector, including venture capitalist Marc Andreessen and former Trump adviser David Sacks, express opposition to mandatory requirements. Sacks in March resigned from his position as Trump’s lead AI official and is currently co-chairing the president’s technology advisory committee. Trump’s AI policies in his second term have predominantly mirrored the viewpoints of the technology sector. A spokesperson for the White House characterised any discourse regarding the specifics of AI policy as “speculation.” The order emerged from the collaborative efforts of White House Chief of Staff Susie Wiles, White House Science & Technology Adviser Michael Kratsios, Wiles’ deputy Walker Barrett, and Cairncross over the past month, incorporating insights from AI companies, according to a source familiar with the planning.
The balance of power between the two factions of Trump’s supporters has undergone a transformation, influenced by the emergence of advanced AI systems, notably Mythos and OpenAI’s GPT-5.5-Cyber. The companies caution that the new models may significantly enhance the sophistication of cyberattacks, although certain cybersecurity executives have suggested that these concerns may be exaggerated. Mythos’ arrival incited a contest among the president’s supporters to shape his response. The outcome of that debate could significantly influence the AI industry, particularly if the president’s decision delays the rollout of large language models or compels companies to modify model performance in response to safety concerns. Both alternatives have the potential to negatively impact profitability. Historically, Republicans have championed the principles of limited government and resisted regulatory measures. However, there is an increasing momentum among their more outspoken populist base advocating for the establishment of guardrails around artificial intelligence. The populist faction has urged Trump to mandate government approval for “potentially dangerous” AI systems prior to their deployment, as outlined in a letter sent to the White House last Friday. Kremer stated that it is “antithetical” to her political views to advocate for new regulations; however, AI necessitates a different approach. “You cannot rely on the individuals at the helm of these AI companies to prioritise our interests and take the necessary actions to safeguard the American populace,” she stated.
Kremer played a role in organising a rally on January 6, 2021, which took place prior to the Capitol riot. She stated in an interview that she was not part of the thousands of Trump supporters who breached the Capitol that day. Tech executives rank among the president’s most significant political contributors and prominent advocates. Meta’s Mark Zuckerberg, Amazon’s Jeff Bezos, Google’s Sundar Pichai and OpenAI’s Sam Altman occupied prominent positions as he took the oath of office in January 2025. Proponents of the technology sector communicated their desire for the US Department of Commerce’s Center for AI Standards and Innovation to assume a pivotal role in the Trump administration’s approach to advanced AI models. They noted that companies are prepared to collaborate voluntarily with the scientists and cybersecurity experts within that organization. The National Security Agency has participated in administration-wide discussions regarding the appropriate response to Mythos, as indicated by sources. Lawmakers requested Cairncross to collaborate with federal agencies to establish a mechanism for monitoring “sudden frontier AI capability jumps.” Brad Carson stated “The past couple months have served as a massive wake-up call for the kinds of vulnerabilities that AI can create.”
Delaying the release of new AI models for federal review may provide the United States with a temporary edge over its competitors; however, it is unlikely to prevent adversaries from ultimately acquiring the technology, according to Neil Chilson, head of AI Policy at the Abundance Institute, a nonprofit frequently associated with the technology sector. “We need to ensure that we are effectively deploying it and maximising its utility, which includes strengthening our defenses,” Chilson stated. Voluntary federal testing of new AI models has been implemented for several years, with companies like OpenAI and Anthropic presenting their products for evaluation by the Center for AI Standards and Innovation, previously known by another name during the administration of former President Joe Biden. The Commerce Department announced in May that Google, xAI, and Microsoft had consented to submit their AI models for security testing; however, the specifics subsequently vanished from its website. The White House and Commerce Department failed to provide commentary regarding the absence of the details in question.
