Decentralized AI tokens rise after U.S. ban on Anthropic models

A U.S. export-control directive ordering Anthropic to block foreign-national access to its newest models prompted rallies in tokens tied to decentralized AI projects. Venice’s VVV rose about 14% to $16.37 and Morpheus’s MOR climbed roughly 21% to $2.28.

The directive reached Anthropic at 5:21 p.m. ET on Friday. To comply, the company disabled Fable 5 and Mythos 5 for all customers while keeping its other models active. Fable 5, released on June 9, is the first publicly available model in Anthropic’s higher-capability Mythos class.

CoinGecko data showed VVV hit a 24-hour high of $17.66 as trading volume rose nearly 200% to about $130 million. MOR’s gain occurred on roughly $300,000 of volume over the same period.

Anthropic noted regulators were focused on a narrow method that can bypass built-in safeguards and said the problem is shared by other models. The company is working to reverse the directive and reported it would provide more details within 24 hours.

Backers of decentralized AI framed the shutdown as support for censorship-resistant systems. The official Morpheus account posted that “decentralized AI never looked better,” thanked Anthropic CEO Dario Amodei for “free publicity” and offered “condolences” to affected users. Venice founder Erik Voorhees responded, “There's a reason we built Venice.” Former White House official Dean Ball wrote that the order suggests top models may be off-limits to non-U.S. citizens even when those users are inside the United States and that platforms could require citizenship verification.

The directive's scope may extend to foreign-national employees. Several Anthropic staff were born outside the United States, though their citizenship status has not been disclosed.

Venice presents itself as a privacy-focused platform that runs open-source models and grants access through staking the VVV token. Morpheus describes its network as a community project that rewards contributors of compute, code and capital with MOR.

Security experts have warned that more capable models can aid cybersecurity work and also make it easier to find software flaws by automating parts of vulnerability discovery. Anthropic’s Mythos-class models have performed strongly on cybersecurity-related tests.

Anthropic is working to restore access and to clarify the directive with regulators. The immediate token gains reflected investor interest in decentralized alternatives; the longer-term market effect will depend on how quickly the export-control concerns are resolved.