The most striking thing about the standoff that ended on Tuesday is not that it ended, but that it happened at all: a US government blocking a US company from shipping its best product, on national security grounds, to anyone at all.
Anthropic said the Department of Commerce has lifted the export controls that forced it to disable access to its Claude Fable 5 and Mythos 5 models in mid-June.
Fable 5 returns to global users from Wednesday, while access to the more tightly held Mythos 5 has been restored for some approved US organisations.
To understand the fuss around Fable, it helps to separate the two models.
Mythos 5 is the frontier system, the kind of highly capable model that governments worry about in the context of cybersecurity and biological risk, and its distribution has long been restricted.
Fable 5 shares the same underlying model but carries additional safety measures, and its release in June marked the first time Anthropic had put an offering this advanced into the hands of the general public.
That is precisely why the clampdown stung.
The directive on 12 June cited national security authorities and, unusually, extended to any foreign national, including Anthropic's own overseas employees.
A company that had just launched a product it billed as state-of-the-art suddenly could not let much of the world use it.
The commercial logic of the objections was straightforward.
Chinese open-source models have been closing the gap on the leading US systems, matching much of their capability at lower cost.
Executives and investors argued that hobbling Anthropic's rollout simply handed Chinese developers time to catch up, an own goal dressed as a security measure.
The politics ran underneath all of it.
Chief executive Dario Amodei has been a target of the Trump administration, both for his outspoken stance on AI safety and his support for Kamala Harris in 2024.
Tellingly, co-founder Tom Brown reportedly took over the negotiations, and the Commerce Secretary's letter clearing Mythos 5 was addressed to Brown, not Amodei.
For a company widely expected to pursue a public listing, the resolution matters well beyond the immediate product calendar.
An export regime that can switch off a flagship model overnight is exactly the kind of risk that gives IPO investors pause, because it strikes at the predictability of future revenue.
Removing that overhang and demonstrating that Anthropic can negotiate its way back to market restores a cleaner growth narrative just as the window for listing is being debated.
It also signals to prospective backers that the company retains commercial access to international markets, the same markets that underpin its valuation.
The reprieve is not unconditional.
Mythos 5 access is being widened only gradually, through trusted partners and the Glasswing security programme, and the broader executive order asking developers to submit models for government review before release remains in force.
Anthropic has cleared this hurdle, but the machinery that created it is still very much in place.
