Какво трябва да знаем за забраната на САЩ за Anthropic

The Trump administration has blocked foreign users from accessing Anthropic’s newest AI models, Claude Fable 5 and Mythos 5, citing national security concerns over “jailbreaking.” This move by the U.S. Department of Commerce forced the San Francisco company to shut down global access entirely, just as it filed for a public offering.

Washington officials stated that the restrictions are necessary to prevent the bypassing of AI safety rules. Anthropic described these jailbreaking risks as insignificant and exaggerated, noting that competing platforms face similar issues.

Why did the U.S. government block Claude Fable 5 and Mythos 5?

The U.S. Department of Commerce effectively banned foreign citizens from using the models, including some of Anthropic’s own employees. This forced the company to stop global access to ensure compliance with the new export restrictions.

The timing of the ban coincides with Anthropic’s recent filing to begin public trading this autumn. The company had intended to raise tens of billions of dollars from investors through this process.

Did You Know? According to Clemens Füst, president of the German ifo institute, the U.S. controls 75% of global AI infrastructure, while China controls 15% and Europe controls less than 5%.

How does this affect global AI competition?

A group of over 170 tech executives warned in an open letter that these restrictions “threaten America’s leadership in artificial intelligence.” They argued that the ban deprives cybersecurity specialists of essential tools while China’s capabilities continue to grow.

At a G7 summit in France, Indian Prime Minister Narendra Modi called for “broad and inclusive” access to U.S. AI models. The United Kingdom requested an exception to the ban, but Washington denied the request.

French President Emmanuel Macron warned that the U.S. ability to “pull the switch” provides further reason for the European Union to secure its own AI sovereignty. He stated the EU would not buy models from companies that could revoke access overnight.

Expert Insight: Samantha Carter notes that targeting a domestic firm like Anthropic with export controls creates a precarious precedent. By treating a U.S. company as a vehicle for export restriction, the administration may inadvertently signal to global investors that AI assets are subject to sudden regulatory volatility, potentially chilling the very innovation the U.S. seeks to protect.

What is the conflict between Anthropic and Washington?

Anthropic has previously clashed with the U.S. Department of Defense. The company refused to remove restrictions preventing its AI models from being used for autonomous weapons systems or the mass surveillance of American citizens.

OpenAI strikes deal with Defense Department after Trump bans Anthropic

The company utilizes “Constitutional AI,” a safety-first approach where models are trained to self-critique based on written principles. However, University of Washington computer science professor Pedro Domingos claims Anthropic acts as a “self-appointed moral authority.”

Domingos told DV that Anthropic believes AI is a “mortal danger” and that they should “govern the world.” He suggested that the company’s government friction stems from Anthropic attempting to “appropriate” government functions.

What may happen next?

Anthropic CEO Dario Amodei has urged G7 leaders to prioritize international cooperation over unilateral actions. He encouraged leaders to “resist the temptation to disconnect.”

Risto Wook of the Future of Life institute suggested that Washington may need to introduce clearer regulations similar to those the EU will implement in August. Wook characterized the current U.S. actions as “hasty and uninformed.”

Analysts suggest this precedent could potentially undermine investor confidence or stifle innovation within the U.S. tech sector if more domestic companies are targeted by export controls.

Frequently Asked Questions

Which AI models were restricted by the Trump administration?
The restrictions target Anthropic’s latest updates, specifically Claude Fable 5 and the more powerful Mythos 5.

Why did Anthropic stop global access to its tools?
The company stopped global access after the U.S. Department of Commerce banned foreign citizens, including some Anthropic employees, from using the models.

How did European leaders respond to the ban?
French President Emmanuel Macron stated the ban justifies the EU’s pursuit of AI sovereignty, warning that Europe cannot rely on companies that can suddenly cut off access.

How should nations balance national security with the global distribution of AI tools?