The first-ever use of export control powers against a major AI lab is sending shockwaves through tech and crypto markets alike
The US government just did something it has never done before: ordered a leading AI company to shut off access to its most advanced models worldwide. Commerce Secretary Howard Lutnick sent a formal letter to Anthropic CEO Dario Amodei on June 12, directing the company to immediately suspend exports of its latest AI systems, Fable 5 and Mythos 5.
The legal basis is the 2018 Export Control Reform Act, a law originally designed to keep sensitive technologies out of the hands of foreign adversaries. Until now, no administration had wielded it against a commercial AI lab’s deployed products. That precedent just changed.
What happened and why it matters #
Lutnick’s letter cited national security concerns, specifically the risk that foreign military and intelligence services, with China and Russia named as primary targets, could exploit Anthropic’s frontier models. The directive wasn’t a suggestion. Anthropic complied by revoking global access to Fable 5 and Mythos 5, meaning both domestic and international customers lost access to the models abruptly.
The trigger appears to have been security vulnerabilities discovered by Amazon’s research team. Amazon is Anthropic’s largest investor and cloud infrastructure partner.
The tech industry’s response was immediate and hostile. More than 50 cybersecurity leaders and tech CEOs sent letters to Secretary Lutnick, calling the restrictions “dangerous.”
The crypto market’s answer: decentralize everything #
According to CoinGecko, the total market capitalization of decentralized AI tokens climbed to $24.3 billion, jumping 6% in a single day and 12% over the trailing week.
If centralized AI labs can be forced to revoke access overnight, then platforms built on decentralized networks, where no single entity holds the kill switch, suddenly look less like speculative experiments and more like strategic hedges.
Background: a regulatory landscape in flux #
The Lutnick directive didn’t emerge from a vacuum. Washington has been tightening the screws on AI-related exports for years, particularly targeting advanced semiconductor chips bound for China. But those restrictions targeted hardware, the physical components needed to train AI models. This is the first time the government has gone after the software itself, and specifically against models that were already deployed and in active use.
The 2018 Export Control Reform Act gave the Commerce Department broad authority over “emerging and foundational technologies” deemed critical to national security.
Anthropic was founded by former OpenAI researchers who believed the industry wasn’t taking AI risks seriously enough. Being forced to disable your own models by a government letter is a particularly bitter irony for a company whose entire brand is built on responsible development.
What this means for investors #
If the Commerce Department can order Anthropic to disable models today, it can do the same to OpenAI, Google DeepMind, or any other US-based AI lab tomorrow. That regulatory overhang now hangs over every centralized AI company. The 50-plus tech leaders who called these restrictions dangerous aren’t being dramatic. If your API provider can be forced to pull the plug without warning, your business continuity plans need a serious rethink.
The $24.3 billion decentralized AI token market is still relatively small in the broader crypto landscape. A 6% daily move on that base is meaningful but not transformative. The real question is whether this event marks the beginning of sustained capital rotation into decentralized AI infrastructure, or whether it’s a knee-jerk reaction that fades once headlines move on.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our