President Trump signed a new executive order on artificial intelligence on 2 June 2026. The order is titled “Promoting Advanced Artificial Intelligence Innovation and Security.” It focuses on cybersecurity and national security rather than privacy or consumer protection. The White House described its goal as strengthening defences while supporting innovation.
The order covers government agencies, AI developers, and critical infrastructure operators. It targets the risks posed by the most capable AI systems. The administration chose a voluntary framework over a compulsory licensing regime.
A Voluntary Framework With Three Core Structures
Legal analysis published by JD Supra confirmed the order creates no mandatory approval process. Developers face no requirement to obtain clearance before releasing new models. The framework instead builds three linked structures.
The first is an AI cybersecurity clearinghouse. The second is a classified process for identifying what the order calls “covered frontier models.” The third is a 30-day pre-release review window. Developers may use this window before wider deployment. Participation remains optional.
The White House fact sheet states the framework covers national security systems and civilian federal information systems. The order establishes a classified benchmarking process. Officials use it to determine when a model qualifies as a covered frontier model.
The Role of the NSA and Federal Agencies
The National Security Agency takes a leading role under the order. It works alongside senior officials in cyber and science policy. Mayer Brown and Venable both noted that the covered frontier model classification targets systems capable of finding or exploiting software weaknesses at scale.
The Cybersecurity and Infrastructure Security Agency receives a directive to strengthen protections. It must share AI-enabled security tools more widely across government. The Treasury Department must stand up the clearinghouse. Treasury will work with industry to identify and patch software vulnerabilities.
The Attorney General receives a separate instruction. The order directs the Attorney General to prioritise criminal enforcement against anyone who uses AI for hacking or data theft. The full executive order sets out the specific agency responsibilities in detail.
Federal agencies also receive a broader push. The order directs them to use AI more aggressively in defence and security operations. This signals a deliberate choice to embed AI into federal cyber capabilities at scale.
Privacy and Algorithmic Concerns Remain Outside the Order’s Scope
The order introduces no new privacy rights. It creates no new data protection obligations or consent rules. It says nothing about algorithmic bias, workplace impacts, or data subject rights.
Organisations handling large volumes of personal data still operate under the existing framework. That framework includes the patchwork of state privacy laws, sector-specific rules, and enforcement by agencies such as the Federal Trade Commission. The executive order does nothing to alter or consolidate those obligations.
JD Supra’s legal commentary noted this gap directly. The absence of privacy provisions is a deliberate feature of the order’s design. Federal policy under this administration treats advanced AI as a security matter. Privacy concerns remain the territory of existing law and state-level action.
What the Order Signals About Federal AI Policy
The order is the clearest federal move yet into AI oversight under this administration. It does so through a security lens. The decision to keep the framework voluntary reflects a broader preference. The administration wants to avoid regulatory barriers that could slow AI development.
The clearinghouse model and the frontier model classification process both show a preference for coordination over compulsion. Industry involvement is built into the structure. Government and private sector actors share responsibility for identifying and addressing vulnerabilities.
For AI developers and critical infrastructure operators, the order creates new engagement channels with federal agencies. For privacy advocates and consumer protection groups, the order leaves the existing landscape unchanged. The direction of federal AI policy is now security-first. Other concerns remain for other forums.
Subscribe
Sign-up to receive our newsletter
