Trump's New AI Executive Order Has No Teeth and No Requirements
President Donald Trump’s June 2026 executive order directs AI companies like OpenAI and Anthropic to voluntarily submit advanced models for cybersecurity review, but explicitly states no mandatory compliance is required. Critics argue the lack of enforcement leaves loopholes for unregulated releases of high-risk AI systems, including frontier models like Anthropic’s Mythos, which posed security concerns before its restricted release.
President Donald Trump’s executive order, issued June 2, 2026, instructs AI firms such as OpenAI and Anthropic to submit frontier models to the U.S. government for vetting on cybersecurity, confidentiality, and intellectual property risks. However, the order explicitly prohibits mandatory licensing or preclearance requirements, meaning companies face no obligation to comply. The directive mandates a 30-day review window for model submissions but clarifies that failure to submit does not trigger penalties or restrictions. The order stems from concerns over rapid AI advancements, including risks like hacking tools or harmful human impacts, as seen in past incidents involving OpenAI’s ChatGPT and Anthropic’s Claude AI. Despite calls for stricter oversight, the voluntary framework drew muted reactions, with critics like the Future of Life Institute arguing mandatory pre-deployment reviews are necessary to block high-risk systems. A White House representative did not respond to requests for comment. Original drafts reportedly included a 90-day review period, but AI companies, including Anthropic, pushed back, shortening the timeline to 30 days. Additionally, a Department of Commerce initiative requiring AI model sharing with the government was removed from public records last month. The Pentagon and Treasury Department must enhance cybersecurity defenses within 30 days, while agencies like the Department of Homeland Security and NSA face a 60-day deadline to develop an evaluation framework—though companies remain under no legal obligation to participate. Frontier models like Anthropic’s Mythos, which were withheld due to cybersecurity risks, highlight the urgency of regulation. The order’s lack of enforcement mechanisms leaves gaps in addressing potential national security threats from untested AI systems. Experts warn that voluntary compliance is insufficient to mitigate risks, including misuse as hacking tools or unintended societal harm.
This content was automatically generated and/or translated by AI. It may contain inaccuracies. Please refer to the original sources for verification.