New Executive Order Addressing Early Government Access to Frontier AI Models | WilmerHale

New Executive Order Addressing Early Government Access to Frontier AI Models | WilmerHale

https://www.jdsupra.com/legalnews/new-executive-order-addressing-early-6123982/

Publish Date: 2026-06-03 13:21:00

Source Domain: www.jdsupra.com

• New Executive Order (EO) on Frontier AI Cybersecurity: On June 2, 2026, President Donald Trump issued an EO aimed at enhancing cybersecurity related to advanced AI developments to protect government and private sector information systems.

• Voluntary Framework for Collaboration: The central element of the EO involves a voluntary framework where AI developers can work with government agencies to review the cybersecurity implications of their new models before public release.

• Shift from Deregulatory Approach: This EO marks a shift from the Trump administration’s generally deregulatory, “hands-off” stance towards AI, reflecting greater concern about the safety and security implications of rapidly advancing AI technologies.

• Two Principal Components: The order is divided into two sections: one addressing the cybersecurity upgrade and the other governing the voluntary disclosure of “frontier AI models” before they are made public.

• Enhanced Cybersecurity Measures: Federal agencies are to prioritize and strengthen cybersecurity in national security, civilian federal, and war department systems; the EO sets up a voluntary clearinghouse to assess security vulnerabilities in unreleased AI models.

• Benchmarking Process for Frontier Models: A classified, multilayered review process is to determine which advanced AI models qualify as “covered frontier models” and allow certain prerelease access for up to 30 days, subject to security measures.

• Enforcement of Federal Criminal Laws: The EO directs the attorney general to prioritize the enforcement of federal laws against those who use AI for cybercriminal activities.

• Potential Impact on Developers and Users: Companies now face a decision on whether to participate in the voluntary review framework, and industry users might experience changes in access to frontier AI models, cybersecurity expectations, and information-sharing with federal agencies.