bbieron@platformeconomyinsights.com

President Trump Issues Pared Back Executive Order on AI Security Reviews

Jun 13, 2026

Report from Bloomberg
In Brief – President Donald Trump has signed an executive order establishing a voluntary framework to address potential new cybersecurity risks posed by advanced artificial intelligence systems. The directive reflects the administration’s preference for encouraging cooperation between government and industry rather than imposing regulations. Under the order, AI developers may voluntarily provide the federal government with access to frontier AI models for up to 30 days before public release. The policy does not require AI model “safety” testing, licensing, or government approval. Instead, it creates mechanisms for information sharing between AI companies and federal agencies, including the Treasury Department, Pentagon, and National Security Agency (NSA). The NSA, working with other agencies and AI labs, will help establish standards for determining which AI models qualify for government review. In May, the President rejected a plan that included a 90-day review period.

Context – The order follows weeks of reportedly intense debate inside the administration following the announcement by Anthropic that their Mythos model had unprecedented capabilities to find and exploit cybersecurity vulnerabilities. The company privately shared the system with a range of industry and government organizations to help them identify and address cyber vulnerabilities. The AI policy of the second Trump Administration has been aggressively pro-investment and innovation, and anti-regulatory, especially rejecting “handwringing about safety” or governance of AI to address “ideological, risk-focused obsessions, such as climate or equity.” But there is a cadre of high-profile conservative activists, and Republicans in Congress, who harshly criticize AI, Big Tech and its leaders, and call for AI regulation. The administration is clearly torn over whether the best way to deal with “science fiction” style AI risks is still to wave them off, especially as polling shows that Republican and Democratic voters are equally concerned about AI and states are moving forward with regulatory proposals. The AI industry should definitely prepare for a major regulatory push under a unified Democratic federal government.

View By Monthly
Latest Blog
Snap Settles Latest Social Media Addiction Teen Bellwether Lawsuit

Report from The Hill In Brief – Snapchat has settled claims from two teen plaintiffs alleging that the platform was intentionally designed to be addictive to young users, ending a high-profile Los Angeles trial two days before opening arguments. The plaintiffs, from...

EU Commission Wants Frontex to Monitor Online Immigration Content

Report from The Next Web In Brief – The European Commission has proposed giving Frontex, the EU’s border agency, authority to monitor social media posts about migration and flag problematic content. The proposal, included in a broader Frontex overhaul, would establish...

DC Attorney General Joins Thirteen States Suing TikTok Over Safety

Report from Fox 5 In Brief – The District of Columbia has joined the litigation efforts of 13 states suing TikTok, accusing the social media platform of harming children through intentionally addictive design, deceptive claims about user safety, and other violations...

Social Media Giants Lose Challenge Against Ofcom Information Demands

Report from Courthouse News In Brief – The High Court in London has rejected legal challenges by Meta, TikTok and X against demands from Britain’s online safety regulator, Ofcom, for information about how the platforms moderate harmful content. The ruling backs the...

Platform Economy Insights produces a short email four times a week that reviews two top stories with concise analysis. It is the best way to keep on top of the news you should know. Sign up for this free email here.

* indicates required