Trump Asks AI Firms to Share New Models 30 Days Early

AI companies must now give the government a 30-day heads-up on new advanced AI models. This is a change from previous plans that allowed for longer review times.
By Newsroom | Technology, Government |

Government to "Voluntarily" Vet Frontier AI Before Public Release

President Donald Trump on Tuesday signed an executive order that asks artificial intelligence companies to voluntarily provide the federal government with access to their most advanced models up to 30 days before a public release. This measure aims to allow for the assessment of potential cybersecurity risks and protection of critical infrastructure.

Karnataka CM-designate Shivakumar, Siddaramaiah hold high-level meet with Kharge, Rahul over Cabinet plan - The Times of India - 1

The executive order explicitly states that this voluntary pre-release mechanism does not authorize mandatory governmental licensing, preclearance, or permitting requirements. Major AI firms were involved in the development of this directive.

Karnataka CM-designate Shivakumar, Siddaramaiah hold high-level meet with Kharge, Rahul over Cabinet plan - The Times of India - 2

"Common-Sense Approach" to Balancing Innovation and Security

A White House spokesperson characterized the executive order as a "common-sense approach of collaborating with industry to balance innovation and security," and to "cement America’s continued global dominance in AI and cybersecurity." The directive also instructs national security agencies to bolster their cybersecurity defenses, proposing a "cybersecurity clearinghouse."

Read More: Morgan Wallen Piano Incident Searches Show Fashion Brand

Karnataka CM-designate Shivakumar, Siddaramaiah hold high-level meet with Kharge, Rahul over Cabinet plan - The Times of India - 3

The order comes in the wake of growing concerns within government and financial sectors, particularly following the announcement of advanced models like Anthropic’s Mythos. This particular model had reportedly spooked officials and Wall Street, highlighting anxieties surrounding the rapid advancements in AI capabilities.

Karnataka CM-designate Shivakumar, Siddaramaiah hold high-level meet with Kharge, Rahul over Cabinet plan - The Times of India - 4

Industry Input and Revised Timelines

AI companies, facing potential scrutiny and a boom in their fortunes, have played a central role in shaping the administration's stance on AI oversight. The tech industry had reportedly pushed for a shorter review window, with some insiders suggesting a period closer to two weeks. The final order’s 30-day window represents a reduction from an earlier draft that reportedly allowed for up to 90 days of review.

This latest executive order was initially slated for a ceremony on May 20, but was postponed. The earlier version, shelved on May 21, was reportedly a "slimmed-down" version of a previous iteration, which had also included a cybersecurity component but proposed a longer review period.

Read More: DOJ Files New Charges Against Southern Poverty Law Center

Broader Directives and Enforcement

Beyond the pre-release access provision, the executive order directs the Department of Justice to prioritize enforcement against crimes such as AI-assisted hacking and unauthorized access. The Department of Defense is also tasked with prioritizing the cyber defense of its information systems.

The National Security Agency and the Department of Defense are expected to assist in identifying which AI models warrant government scrutiny. The Treasury Department will reportedly focus on identifying vulnerabilities within these AI models.

Background and Context

The current executive order signifies a shift for an administration that has historically shown resistance to regulating artificial intelligence. It arrives as the Pentagon seeks to integrate "frontier AI" into its military operations. This move also follows a previous AI-focused executive order from December, which aimed to prevent states from regulating AI and established a federal taskforce to challenge state-level AI laws. Earlier in May, the Trump administration had reportedly reached an agreement with major tech players, including Microsoft, Google DeepMind, and xAI, to review early versions of their new AI models before public release.

Read More: WhisperX and LLM Use One 3090 GPU with 24GB VRAM

Frequently Asked Questions

Q: What did President Trump order about new AI models on Tuesday?
President Trump signed an order asking AI companies to voluntarily give the government access to their most advanced AI models up to 30 days before they are released to the public. This is to check for cybersecurity risks.
Q: Does this mean the government will approve all new AI models before they are released?
No, the order states this is voluntary and does not give the government power to require licenses or approvals. Companies are asked to share, not forced to get permission.
Q: Why is the government asking for early access to AI models?
The government wants to check for potential cybersecurity risks and protect important systems like critical infrastructure. They also want to make sure America stays a leader in AI and cybersecurity.
Q: Which AI companies were involved in making this order?
Major AI firms were involved in creating this directive. Companies like Microsoft, Google DeepMind, and xAI had reportedly agreed to review early versions of their new AI models before public release.
Q: What happens next with AI development and security?
The government will use the early access to identify risks. The Department of Justice will also focus on prosecuting AI-assisted crimes, and the Department of Defense will improve its cyber defenses.