On 2 June 2026, the White House issued the third major AI executive order of the Trump administration's second term: "Promoting Advanced Artificial Intelligence Innovation and Security." The Order follows the day-one revocation of Biden's 2023 AI safety order, the January 2025 AI Action Plan directive, and the December 2025 executive order establishing a national policy framework and AI Litigation Task Force aimed at preempting state AI laws.
What the June 2026 Order does
The June Order has two primary focuses: cybersecurity and voluntary frontier model security.
On cybersecurity, the Order directs multiple federal agencies — including the Department of Homeland Security through CISA, the Department of Defense, and the Treasury — to prioritise cybersecurity of government information systems within 30 days. This includes establishing or expanding AI-enabled defensive tools and facilitating access to cybersecurity services for state and local authorities. The Attorney General is directed to prioritise enforcement of applicable federal criminal laws against AI-driven cybercrime. Federal agencies must accelerate hiring of cybersecurity specialists within 60 days.
On frontier models, the Order establishes a voluntary benchmarking and review framework for the secure development and deployment of frontier AI models. The framework is developed in consultation with AI developers and is not a mandatory licensing or safety testing regime. It is closer in character to the UK's voluntary AI Safety Institute model than to a binding regulatory framework.
What the Order does not do
The June Order does not preempt any additional state AI laws beyond what the December 2025 executive order already covered. It does not create new mandatory obligations on private companies for AI safety, risk management, or disclosure. It does not establish any new federal AI regulatory body or mandate. Congress has still not passed a comprehensive federal AI law; without legislation, the federal AI policy landscape remains an executive-action-only environment that can shift with administrations.
The three-EO arc
Reading the three executive orders together provides a clearer picture of the administration's posture. The January 2025 actions revoked Biden-era safety requirements and established a pro-innovation framing. The December 2025 order created the policy architecture for federal preemption of "onerous" state AI laws through the AI Litigation Task Force. The June 2026 order adds the security dimension: advancing US AI capability while protecting against AI-enabled threats to federal systems and critical infrastructure. The consistent thread is that AI is a national security and economic competitiveness asset, not a consumer protection or civil rights challenge — which shapes which risks get regulatory attention and which do not.
What this means for organisations
For organisations primarily operating in Australia, the EU, or Singapore, the June EO has limited direct impact. Its cybersecurity mandates apply to federal agencies, not private companies. The voluntary frontier model framework affects developers of frontier systems, not enterprise deployers.
For organisations with US federal contracting, the cybersecurity mandates create indirect obligations — federal agencies will impose new AI security requirements on their vendors and contractors as they operationalise the Order. Organisations in federal supply chains should monitor contracting requirements from CISA, the Department of Defense, and Treasury for AI-related security provisions in the coming 90 days.
The broader US compliance picture has not changed: existing state laws remain in effect and enforceable. Connecticut SB 5 (signed 27 May 2026) and Colorado SB 26-189 (signed 14 May 2026) are binding law; neither is affected by the June EO. The AI Litigation Task Force from the December 2025 order remains the federal government's primary tool for challenging state AI laws, and it has not yet invalidated any state statute — court proceedings take time.
Primary sources: Lathrop GPM — New EO Signals Evolving Federal Approach to AI (June 2026) | Inside Privacy — White House Releases June 2026 AI EO