Reversal of Export Controls
The Commerce Department lifted restrictions on all of Anthropic’s artificial intelligence models on Tuesday, allowing the company to bring its most powerful technologies back online. Consequently, Commerce Secretary Howard Lutnick informed the firm that it no longer needs a license for exports or in-country transfers of its Claude Mythos and Claude Fable models. This decision officially reverses an earlier order issued on June 12, which had suspended access for all foreign nationals due to national security concerns. As a result, this regulatory move de-escalates a growing feud with the Trump administration and paves the way for Anthropic to return to normal business operations.
New Security Agreements and Safeguards
Following weeks of intense deliberations, the government lifted these strict controls because Anthropic agreed to implement comprehensive safety measures. Specifically, the company committed to proactively detecting and addressing security risks within its advanced AI models. Furthermore, Anthropic will work closely with the United States government on safety protocols for future models and promptly inform officials of any malicious activity. However, Lutnick explicitly warned that the Commerce Department reserves the right to re-evaluate these decisions if the company fails to adhere to its new commitments. Following these agreements, Anthropic announced it would begin restoring global access to its models on Wednesday.
Shifting Government Stance on AI
This recent regulatory action highlights a significant shift in the Trump administration’s approach to the rapidly growing artificial intelligence industry. Previously, President Trump took a hands-off approach, arguing that the tech sector needed to grow unfettered to effectively compete with China. Nevertheless, he recently signed an executive order asking tech companies to voluntarily give the government oversight of new models before public release. In addition, these recent export restrictions marked the second time the administration targeted Anthropic; earlier in March, the Pentagon controversially labeled the company a supply chain risk, prompting a swift lawsuit from the San Francisco-based firm.
Wider Industry Implications
The administration has also aggressively moved to regulate other major Silicon Valley developers under this new oversight framework. For instance, when OpenAI recently unveiled its GPT-5.6 Sol technology, it agreed to share the system exclusively with companies explicitly approved by the government. Meanwhile, Meta remains the only major American developer that has not reached a voluntary sharing agreement and was heavily pressed by the administration last week to begin submitting its models. Ultimately, White House Chief of Staff Susie Wiles expressed gratitude to companies cooperating with the executive order, emphasizing a shared national priority of deploying the best technology safely and quickly.
Reference
Frenkel, S., & Swanson, A. (2026, junio 30). U.S. lifts restrictions on Anthropic’s AI models. The New York Times. https://www.nytimes.com/2026/06/30/technology/us-lifts-restrictions-anthropic.html