Howard Lutnick Reverses Export Freeze That Locked Foreign Users Out of Claude Fable 5 and Mythos 5

The Trump administration has abruptly rolled back an unprecedented global lockdown on cutting-edge artificial intelligence systems following intensive, behind-the-scenes negotiations in Washington. US Commerce Secretary Howard Lutnick confirmed the lifting of emergency export bans that had completely pulled the plug on the world's most advanced digital models over national security concerns.

The sudden policy reversal follows a tense two-week standoff sparked by fears that foreign adversaries could exploit the technology to launch sophisticated cyber attacks.

A high-stakes clash between the Trump administration and Anthropic has come to an end following Tuesday's announcement that the US Department of Commerce had cleared the firm's Claude Fable 5 and Mythos 5 models for export.

'We're grateful to our users for their patience, and to everyone who worked with us on redeploying the models,' Anthropic said in a post on X.

Why Anthropic Suspended Its AI Models

To comply with a government trade directive invoking 'national security authorities', Anthropic deactivated its Fable 5 and Mythos 5 platforms in the middle of June. According to the tech firm, federal officials ordered an immediate freeze on usage 'by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.'

According to an official company announcement, international users can access Fable 5 across Claude.ai, Claude Code and the wider Claude platform once more from Wednesday, 1 July.

Corporate statements confirmed that the AI model will account for up to half of the weekly usage quotas until 7 July for those on Pro, Max and Team tiers, alongside certain corporate accounts.

Following federal clearance received on 26 June, Anthropic confirmed that select American entities have regained access to Mythos 5. The firm plans to work with US officials to broaden this availability for additional domestic and overseas partners via Glasswing, its dedicated cyber-defence project that opens sophisticated AI systems to vetted organisations for vulnerability screening.

Plans are also underway at Anthropic to bring Fable 5 back to Amazon Web Services, Google Cloud and Microsoft Foundry at the earliest opportunity.

China Competition Raises Pressure

Washington's aggressive pushback against Anthropic played out alongside the rapid ascent of open-source software from China, which is matching top-tier American systems in performance at a fraction of the cost. As the Trump administration stalled the release of Anthropic's cutting-edge software, venture capitalists and tech leaders warned that overseas rivals were being handed a crucial opportunity to close the technological gap.

This latest update follows shortly after US Commerce Secretary Howard Lutnick authorised Anthropic to deploy Mythos 5 to a restricted group of government departments and private firms. According to correspondence obtained by CNBC, Lutnick said he was satisfied that 'appropriate safeguards' had been established, allowing vetted 'trusted partners' to use the system.

'Over the past two weeks, we have worked closely with Anthropic to analyze and approve Fable 5 to ensure alignment across the US Government and strengthen America's leadership in AI,' Lutnick wrote in a post on X on Tuesday.

Shortly before the trade restrictions came into force, Anthropic had unveiled the systems while showcasing their record-breaking performance across multiple industry metrics. The rollout of Fable 5 was especially significant, representing the tech firm's inaugural release of a commercially available model with that level of sophistication.

Behind the Washington Negotiations

Following a period of persistent friction with the White House throughout the year, Anthropic travelled to Washington, DC, to hold discussions with the Trump administration after being served with an export control order on 12 June. Since then, the firm has maintained a guarded stance, providing very little information publicly about the progress of those negotiations.

Reports indicate that Anthropic co-founder Tom Brown spearheaded the discussions with the Trump administration, stepping into the negotiating role instead of Anthropic CEO Dario Amodei. Amodei had drawn intense scrutiny from White House officials due to his public stance on artificial intelligence safety and his endorsement of Kamala Harris during the 2024 presidential campaign. Reflecting this shift, Lutnick sent Friday's formal correspondence directly to Brown rather than Amodei.

Industry Faces Fresh Uncertainty

The executive order issued in early June required developers to voluntarily submit models for capability checks before public release, giving federal agencies 60 days to establish formal frameworks.

The approach has sparked widespread confusion across the industry, particularly after AI czar David Sacks vacated his post, leaving tech leaders uncertain about who now influences administration policy.

While OpenAI complied with the directive during Friday's launch of three new systems, including GPT-5.6, by restricting initial access to vetted partners, the firm openly warned against making such pre-release government reviews the permanent standard.