OpenAI said Friday, June 5, 2026, that it will comply with President Donald Trump’s new executive order asking artificial-intelligence companies to provide their most powerful AI models to the federal government for testing before they are released to the public.

Speaking to CNBC on the sidelines of the SXSW London festival, George Osborne, OpenAI’s Head of Countries, said the company would participate in the voluntary program, arguing that democratic governments have a legitimate role in shaping how advanced AI systems are developed and deployed.

The decision marks one of the clearest signs yet that Washington and Silicon Valley are moving toward a more structured relationship as AI systems become increasingly powerful and influential.

What the Order Does

Trump signed the executive order on June 2, directing federal agencies to establish a voluntary framework under which developers can submit their most advanced AI systems for government testing up to 30 days before public release.

The purpose is to evaluate advanced cyber capabilities and determine whether a model qualifies as a “covered frontier model” — a designation reserved for the most powerful AI systems that could present significant national-security or cybersecurity implications.

Importantly, the order does not create a licensing system or require government approval before a company can launch a model. Participation remains voluntary, reflecting the administration’s effort to balance innovation with security concerns.

The order directs agencies including the Department of Defense, the Department of the Treasury, and the Cybersecurity and Infrastructure Security Agency (CISA) to strengthen cyber defenses and establish testing procedures for advanced AI systems.

Why Washington Changed Course

The shift follows growing concern that frontier AI models may be capable of identifying and exploiting cybersecurity vulnerabilities at unprecedented speed.

Those concerns intensified after Anthropic announced in April that it would limit the release of its Mythos Preview model because of its ability to discover and exploit software vulnerabilities.

Security experts have increasingly warned that future AI systems could dramatically accelerate cyberattacks, identify previously unknown weaknesses, and automate sophisticated offensive operations. Government officials argue that early testing could help identify major risks before powerful systems are released widely.

The move represents a notable evolution for an administration that had previously favored a lighter-touch approach to AI oversight.

The Business Impact

The order could have significant implications for the rapidly growing AI industry.

A 30-day government review introduces a new step between completing a model and releasing it to customers. While the review remains voluntary, companies that participate may gain credibility with governments, enterprise customers, and investors concerned about safety and security.

The framework may also strengthen the competitive position of larger AI developers such as OpenAI, Anthropic, Google DeepMind, Microsoft, and xAI, all of which have the resources to absorb additional compliance and testing requirements.

Smaller developers and foreign competitors could face greater challenges if government testing eventually becomes an industry expectation.

The policy could also create additional hurdles for emerging open-source competitors, including Chinese developers such as DeepSeek and Qwen, whose models have rapidly narrowed the performance gap with leading U.S. systems.

In effect, the companies best positioned to comply with expanded safety reviews may gain an advantage as governments, corporations, and regulators increasingly prioritize trust and security.

Industry Response

OpenAI is not alone in embracing government oversight.

Google DeepMind, Microsoft, and xAI previously agreed to submit advanced models for review through arrangements with the U.S. Center for AI Standards and Innovation, while OpenAI and Anthropic entered similar agreements in 2024.

Reaction to Trump’s executive order has been broadly supportive.

Microsoft President Brad Smith called the measure an important step toward balancing innovation and public safety. Anthropic described it as a meaningful move to strengthen America’s leadership in artificial intelligence while reducing risks associated with increasingly powerful systems.

OpenAI has gone even further than the administration’s proposal.

In its policy paper, “Democratic Governance of Frontier AI,” the company advocated for a national AI safety framework that includes mandatory testing of the most advanced systems, independent audits, and whistleblower protections. OpenAI argued that decisions regarding frontier AI safety should ultimately be guided by democratic governments rather than individual technology companies.

What Comes Next

The success of the new framework depends entirely on continued cooperation from the companies building the world’s most powerful AI systems.

National-security officials want early visibility into frontier models because of their potential cybersecurity implications. Technology companies, meanwhile, want to maintain public trust while avoiding regulations that could slow innovation.

For now, the arrangement remains voluntary.

Whether that cooperation continues as AI systems become more powerful — and whether future administrations seek stronger oversight — may determine how artificial intelligence is governed for years to come.

Sources: White House Executive Order on Frontier AI Model Testing (June 2, 2026); CNBC interview with George Osborne at SXSW London (June 5, 2026); OpenAI policy paper Democratic Governance of Frontier AI; U.S. Center for AI Standards and Innovation agreements.

JBizNews Desk — Technology

© JBizNews.com All Rights Reserved. Reproduction or distribution without written permission is prohibited.