The suspension is over. Anthropic’s redeployment announcement confirms that US export controls on Claude Fable 5 and Claude Mythos 5, applied on June 12, 2026, have been lifted as of June 30. Fable 5 returned to global availability on July 1 across the Claude Platform, Claude.ai, Claude Code, and Claude Cowork. After nearly three weeks of disrupted access, and a suspension that caught every user, domestic and international, in its net, the model is back.
The July 7 deadline is the most operationally urgent detail. For Pro, Max, Team, and select Enterprise plans, Fable 5 is included at no additional charge for up to 50% of weekly usage limits through July 7. After that date, access continues, but only via usage credits. Teams that assumed Fable 5 would remain a flat-cost included model need to revisit that assumption immediately. The 50% weekly usage cap during the included window is also a constraint the original Wire reporting omitted; at high volumes, that cap matters.
Why it matters
This is the resolution of the most disruptive AI access event of 2026 so far. When the June 12 export controls landed, Anthropic couldn’t verify user nationality in real time, so it suspended access for everyone, not just foreign nationals. Enterprise teams that had built workflows on Fable 5 lost access overnight with no warning. The redeployment restores that access, but the conditions have changed: the included window is time-limited, platform re-enablement on AWS, Google Cloud, and Microsoft Foundry is still pending per Anthropic’s announcement, and Mythos 5 is not fully restored.
On Mythos 5: Anthropic describes the restored access as covering “a set of US organizations,” following US government approval on June 26. That framing is deliberately narrow. Don’t assume Mythos 5 is broadly available, it isn’t. The criteria for which US organizations qualify haven’t been made public.
Who This Affects
Context
The June 12 export controls were a federal directive that Anthropic had no advance notice to prepare for. The company’s own announcement states the order “took effect immediately” with “no reliable way to verify nationality in real-time”, which is why the suspension was universal rather than targeted. That context matters for enterprise risk planning: even a well-resourced lab can lose access to its flagship models overnight if the government issues an immediate order. This isn’t a hypothetical scenario anymore. It happened. It resolved. And the pattern of restricted frontier AI access is the subject of deeper analysis in as of publication’s synthesis piece.
Anthropic has stated the redeployment includes system-level protections against malicious cybersecurity use, though the company didn’t detail the specific implementation in its public announcement.
What to watch
Three things. First, the July 7 cutoff: teams on Pro, Max, Team, or select Enterprise plans need to decide now whether to activate usage credits before that date or plan for a transition. Second, platform re-enablement: AWS, Google Cloud, and Microsoft Foundry integrations are pending, if your team uses Fable 5 through one of those distributions, you’re still waiting. Third, Mythos 5 access criteria: the government approved a “set of US organizations” without public criteria. If your organization believes it qualifies, the next step is engaging Anthropic directly.
TJS synthesis
Wait for platform re-enablement before rebuilding workflows that depend on AWS Bedrock, Google Cloud Vertex, or Azure AI integrations. The Claude Platform and Claude.ai are live now, but the major cloud distributions, where most enterprise production workloads actually run, aren’t confirmed back yet. Rebuilding on the assumption of full availability and then hitting a second access gap is a preventable outcome. Confirm your distribution channel is restored before you recommit engineering resources.