US AI export controls on Anthropic may trigger sector rotation, UBS says

US export controls on Anthropic's frontier AI models could spark a rotation from semiconductor to software stocks, according to UBS.

The US government on June 12 ordered Anthropic to block foreign nationals from accessing its two most advanced AI models — Fable 5 and Mythos 5 — citing a jailbreak method that could bypass cybersecurity safeguards. The export control, issued by the Commerce Department's Bureau of Industry and Security, forced Anthropic to disable the models globally just three days after Fable 5's public release.

"The administration issued this export control reluctantly after Anthropic CEO Dario Amodei refused to fix the jailbreak or de-deploy the model," David Sacks, co-chair of Trump's Council of Advisors on Science and Technology, wrote on social media. "The hope now is that Anthropic remediates the safety issue, the control is lifted, and Fable goes back into general release."

UBS analysts said the restrictions could trigger a selloff in semiconductor stocks on reduced demand expectations from constrained AI model deployment, while software stocks may benefit from competitive advantages or increased demand for compliant AI solutions. The regulatory shift could significantly reshape the AI investment landscape, the bank said in a note.

The regulatory backdrop

The dispute between Anthropic and the Trump administration has escalated since early 2025, when the administration accused the company of making "woke AI" and called Amodei an "ideological lunatic." Tensions sharpened after Anthropic declined to let the Pentagon use its models for domestic surveillance and fully autonomous weapons systems. The Department of Defense responded by threatening to designate Anthropic a "supply chain risk" — a classification that would have required military contractors to sever ties.

Amazon CEO Andy Jassy was among tech leaders who raised security concerns about Anthropic's models to senior Trump administration officials before the order, a person familiar with the matter told Reuters. Amazon, which is both a rival to and a significant investor in Anthropic, did not confirm whether it spoke to government officials about the models.

Anthropic said it believes the government became aware of a jailbreak — a method for circumventing the safeguards in Fable 5 that prevent using its most powerful features for nefarious purposes. The safeguards classify user requests as safe or unsafe before passing them to the AI model, redirecting unsafe requests to a less powerful model. The bypass found only "minor" security flaws that other publicly available models can also find, Anthropic said.

Export controls with global reach

Some experts questioned the scope of the restrictions. "This was not well thought-out," said Jimmy Goodrich, a senior fellow at the University of California's Institute for Global Conflict and Cooperation. "It even bans Canadians and Brits employed at Anthropic from doing research and development."

The order came as the administration has moved from an initial hands-off posture on AI to asking developers to share models for review before release — an implicit admission that it does not trust companies to fully evaluate what their own models can do. A survey across 25 countries last year found people are more than twice as concerned about AI as they are excited about it.

The Information reported that the administration is unlikely to force other AI firms to abide by restrictions similar to those placed on Anthropic. The current average US tariff on Chinese goods stands at about 11 percent after the 2025 escalation rounds, providing a precedent for targeted rather than broad-based technology controls.

The opacity of frontier AI models remains a core challenge for regulators. According to Oxford University economist Maximilian Kasy, large language models work much better than they "should" given their billions of internal parameters, likening modern AI development to alchemy — successful through trial and error, not yet grounded in systematic theory. This makes comprehensive evaluation by either companies or governments extremely difficult.

This article is for informational purposes only and does not constitute investment advice.