- Washington has ordered the American company Anthropic to suspend access of “any foreign national” to several of its new artificial intelligence models.
- The administration points to a risk for national security, but for Paris, this affair shows above all that these tools are used as “instruments of power”.
- The announcement also caused a reaction from the heavyweights of the French presidential election, who pleaded for independence from the United States.
Follow the full coverage
Artificial intelligence, this technology that is changing our lives
An unprecedented announcement, which is already shaking the sector and causing reactions even in Europe, particularly in Paris. The American flagship of artificial intelligence Anthropic (nouvelle fenêtre) was forced by its own government, Friday June 12, to cut off access to its most powerful models, three days after their launch. Washington cited a risk for national security, a decision whose shockwave even shook the French political class, causing several presidential candidates to react.
The American administration ordered, under export control, to cut off access to these models for “any foreign national, whether inside or outside the United States”
including “les employés étrangers”
of Anthropic, according to a company press release published Friday. Unable to sort its users, the company announced it had to “abruptly disable”
its two new models for all of its customers. It therefore announced that it would suspend its very restricted version Mythos 5 and its restricted version for the general public, Fable 5, in order to comply with this injunction.
The company “contests” a decision seen as “a misunderstanding”
According to the American media Axios (nouvelle fenêtre)the directive comes from Commerce Secretary Howard Lutnick. He allegedly acted after learning that a company had managed to circumvent the safeguards put in place for these models, known to detect and exploit cybersecurity vulnerabilities with unprecedented speed and acuity. In a new article published Sunday (nouvelle fenêtre)the media specifies that it was the giant Amazon which sent administration officials a report showing that it had managed to circumvent restrictions and access certain parts of the Mythos model.
De son côté, Anthropic a qualifié d’“étroite”
the opening found by the organization which alerted the American administration, whose name it did not reveal, and “minors”
the software flaws that it allowed to identify. “We dispute that the discovery of a potential circumvention justifies the recall of a commercial model deployed to hundreds of millions of people”
also wrote the San Francisco start-up, which wants to see in the matter “a misunderstanding”
. She claims to be working to restore “as soon as possible”
access to these latest models, the others remaining operational.
Read also
‘We won’t let any other nation beat us’: What’s in Trump’s AI action plan
This is not the first time that the company, which has made security a commercial argument and which has long campaigned for public supervision of AI (nouvelle fenêtre)finds itself in direct conflict with the Trump administration. In early March, the Pentagon terminated its contracts with the company, designated as a “supply chain risk”
. Anthropic, whose models were the only ones accredited as defense secret, then took legal action, claiming to have been sanctioned for having refused to allow its AI to be used for mass surveillance or autonomous weapons.
France calls for “European digital sovereignty”
The decision of the American authorities caused a reaction even on the Old Continent. The European Union, which obtained access to Mythos at the beginning of June, estimated that “Europe’s need for technological sovereignty”
was stronger than ever. “We note Anthropic’s statement and are currently evaluating the situation.”
said Thomas Regnier, spokesperson for the European Commission, which this month unveiled measures aimed at reducing the dependence of the 27 member states (nouvelle fenêtre) vis-à-vis the United States and Asia for key technologies, notably AI.
“More than ever, European digital sovereignty is a necessity”
abondé de son côté la minister française déléguée au Numérique, Anne Le Hénanff. “The great powers now consider strategic technologies as instruments of power”
she observed.
“The AI war has already begun”: contenders at the Élysée react
The decision also caused a reaction from the rest of the French political class, and in particular potential or declared candidates for the 2027 presidential election. “The AI war has already begun”
judged the former Prime Minister Gabriel Attal, boss of Renaissance, who estimated that “we cannot rely on others because that makes us vulnerable”
. “We have mastered neither models nor calculation”
from AI, “as essential as electricity or the Internet”
alerted another ex-tenant of Matignon, Édouard Philippe. Urging Europe to “wake up”
the leader of Horizons encouraged to favor “European technology markets”
and to simplify “standards that benefit
les big tech américaines
(nouvelle fenêtre)“
.
Read also
“The greatest risk for the human species”: does artificial intelligence represent a real danger?
On the right, LR candidate Bruno Retailleau called for “do with AI what we did with nuclear power, (…) think of it as part of our sovereignty”
. The former Minister of the Interior and senator even invited the boss of Anthropic, Dario Amodei, to settle in France (nouvelle fenêtre) : “You are welcome in Paris”
he at-the launch.“This sudden decision reminds us that artificial intelligence is already a major subject of national sovereignty”
echoed on X the president of the far-right RN party, Jordan Bardella, urging “accelerate in support”
to the French company Mistral AI and “tout l’écosystème IA”
.
On the left, the leader of LFI, Jean-Luc Mélenchon, considered that the American decision “proves the urgency of being independent and sovereign”
while seeing a “political score-settling”
of the American government with Anthropic,
“supporter of ethical AI”
. “We have entered a new world (…) where we can only count on ourselves first”
commented for his part the boss of the PS, Olivier Faure, also calling for “build a real European power”
.
