In June 2026, the Trump administration significantly expanded federal oversight of frontier AI through a voluntary pre‑release review program, export controls on Anthropic’s advanced models, and a congressional discussion draft proposing a national AI framework.
OpenAI is preparing to deploy its flagship ChatGPT model across the Department of Defense’s GenAI.mil platform, making it available to over 3 million defense personnel with Impact Level 5 certification.
On July 3, 2026, the Dutch government released a comprehensive international AI strategy aimed at strengthening European AI capacity, reducing strategic dependencies, and fostering global cooperation on safe and responsible AI development.
On June 29, 2026, the Council of the European Union approved amendments to the AI Act under the ‘Omnibus VII’ package, simplifying rules, banning non‑consensual intimate deepfakes, and extending compliance timelines for high‑risk AI systems.
The U.S. General Services Administration (GSA) has issued a proposed rule—GSA Regulation 552.239‑7001—imposing stringent data safeguarding requirements on contractors using large language models (LLMs) with government data. The rule mandates strict ownership, automated “eyes‑off” controls, U.S. jurisdiction, incident reporting, and an “Unbiased AI Principles” framework, with public feedback due by August 3, 2026.
A new bipartisan draft—the Great American Artificial Intelligence Act of 2026—seeks to centralize AI regulation at the federal level by preempting state AI laws for three years, establishing a governance office, and mandating transparency and independent audits for frontier AI models.
On June 2, 2026, President Trump issued an executive order establishing a voluntary 30‑day pre‑release review for advanced AI models and directing the creation of an AI cybersecurity clearinghouse to bolster national cyber defenses.
On May 7, 2026, the European Parliament and Council struck a provisional agreement under the “Digital Omnibus on AI” to streamline implementation of the EU AI Act. The deal delays enforcement of high‑risk AI rules, introduces new protections, and extends support for small and mid‑cap firms.
As of May 19, 2026, the federal TAKE IT DOWN Act’s notice‑and‑takedown requirement for non‑consensual intimate AI‑generated imagery is enforceable, marking the first U.S. federal AI content takedown mandate. Platforms must remove such content within 48 hours of notice, with no fair‑use exceptions.
The European Parliament and Council have reached a provisional agreement to simplify the implementation of the EU AI Act, reducing compliance burdens and clarifying oversight responsibilities.