The U.S. government has reopened the door to Anthropic’s most advanced artificial intelligence systems, lifting restrictions that had briefly cut off access to the company’s flagship models on national security grounds. Anthropic plans to start restoring the models on Wednesday, following the Commerce Department’s latest determination.
The decision unwinds a June 12 order that compelled Anthropic to block access to its Fable 5 and Mythos 5 models for foreign nationals and, effectively, a wider pool of users. U.S. officials had been worried that frontier A.I. models might help identify vulnerabilities in software, enable faster cyberattacks, or otherwise generate security threats if released without stronger guardrails.
Soon after the initial clampdown, the Trump administration indicated it was prepared to adjust course. Mythos 5 was authorized for a curated set of trusted U.S. organizations, including selected companies and critical-infrastructure operators, before the broader reversal took shape. That interim step signaled that the White House was attempting to separate high-risk open access from tightly monitored deployments.
With the latest move, Anthropic can again pursue broader commercial use of both Fable 5 and Mythos 5, closing out a short but closely watched confrontation between the company and Washington. During the pause, Anthropic collaborated with federal officials on additional security layers, including enhanced monitoring and safeguards designed to curb potential misuse.
The policy reversal is significant because it illustrates how the U.S. is treating cutting-edge A.I. as a dual-use technology: a driver of innovation and productivity, but also sensitive enough to justify export and access controls. It also highlights how quickly the politics around A.I. can shift as national security, cyber defense, and the race for technological leadership intersect.
For Anthropic, the change restores access to its most capable systems and removes a constraint that had disrupted some customer access and commercial rollout timelines. For the wider A.I. ecosystem, the episode serves as a warning that even leading developers of frontier models may face sudden regulatory moves if Washington decides a system could present unacceptable security risks.
Read Article: India’s UPI Goes Live in Greece With Eurobank – NIPL Partnership, Marking 10-Country Reach
