Trump AI Order Prioritizes Cyber Defense Over Broader Safety

Trump AI strategy emphasizes cyber threats, not model safety

The first major deadlines under President Trump's AI executive order arrive in July and August, and the emerging policy framework centers squarely on cybersecurity and national security rather than the broader AI safety concerns that dominated the previous administration's approach.

The strategy relies on voluntary cooperation from AI developers instead of mandatory safety requirements, even as leading AI labs publicly warn about concerning capabilities in their newest models.

Why it matters

The shift from mandatory oversight to voluntary frameworks represents a fundamental change in how the federal government approaches AI governance. This comes at a moment when companies like OpenAI and Anthropic are themselves raising alarms about their latest systems, creating a potential gap between industry warnings and regulatory response.

Key deadlines and agency responsibilities

By July 2, the Department of Homeland Security must deliver a plan to strengthen cyber defenses for federal information systems and ensure critical infrastructure sectors—hospitals, banks, and utilities—have access to the latest AI models. The Treasury Department will establish an "AI cybersecurity clearinghouse" with industry partners to identify and patch vulnerabilities, while the Office of Management and Budget identifies grant funding for AI vulnerability detection.

The August 1 deadline brings additional requirements. OMB must expand hiring for cybersecurity specialists through the U.S. Tech Force. Treasury, the National Security Agency, and the Cybersecurity and Infrastructure Security Agency will develop a classified benchmarking process to assess advanced cyber capabilities in AI models and determine which entities fall under that assessment framework.

Those same agencies will also design a voluntary framework allowing AI developers to give the government access to models up to 30 days before public release.

Notable absence of Biden-era testing center

The executive order makes no mention of the Center for AI Standards and Innovation within the Commerce Department, which has been testing AI models for safety since the Biden administration. CAISI recently announced new testing agreements but was instructed to remove that announcement shortly after publication as the White House finalized its AI order, according to a person familiar with the matter.

"CAISI's absence is hard to ignore," said Ilona Cohen, chief legal officer at HackerOne and former Obama administration lawyer. "When an administration leaves its flagship AI safety institution out of a major AI order, people are going to ask whether the center's role is being reduced or simply redefined."

A White House spokesperson said the implementation represents "a whole-of-government effort" focused on strengthening cyber and national security while ensuring U.S. leadership in AI innovation. A Commerce Department spokesperson declined to address questions about CAISI's exclusion.

Concerns about implementation capacity

"There's a risk of a gap developing between what the Trump administration AI policy says and believes about policy implementation," said Michael Horowitz, a former Pentagon official and University of Pennsylvania professor. He cited both talent departures from government and the administration's willingness to rapidly shift approaches as factors that "reduce predictability."

Horowitz noted that frontier AI labs are raising legitimate alarms about their capabilities while also having economic incentives to promote their technological advances. "That's all the more reason for the government to have the capacity to evaluate these models," he said.

These details were first reported by Axios.