Trump and Anthropic: American AI Under Pressure

Washington Imposes Restrictions on Anthropic

At the request of the White House, Anthropic has been forced to withdraw its latest artificial intelligence models, Fable 5 and Mythos 5, from the network. This decision was made without notice, and the American company stated that it had little choice after Washington demanded to block access for all foreign nationals, including its own employees. This sudden shutdown highlighted the significant control that the United States exerts over cutting-edge AI technologies.

The Fable 5 and Mythos 5 models were already subject to strict security measures, limiting their use in "high-risk areas." However, the swift and extensive action taken by the Trump administration has intensified concerns about global reliance on the United States for critical technologies. This situation has provided new arguments for politicians, governments, and companies advocating for self-sufficiency in this technology.

International Reactions: Towards Technological Autonomy

In the United Kingdom, Kanishka Narayan, Minister for AI and Online Safety, used this incident to advocate for the development of a national AI capability. Although he did not directly mention Anthropic or the United States, he stated that AI is "the central political issue of our time," emphasizing the importance of addressing this matter as a threat to national sovereignty. He insisted that the UK must decide how technology will shape its economy, security, and sovereignty "before anyone else decides the answer for us."

In France, Gabriel Attal, former Prime Minister and presidential candidate for the Renaissance party, compared the withdrawal of Anthropic's models to the Iranian blockade of the Strait of Hormuz. He described this event as the beginning of an "AI war" and stressed France's vulnerability if it continues to rely on other nations for critical technologies. Attal also stated that this issue will be at the heart of the upcoming French presidential elections. Le Monde reported similar alarm across the French political spectrum, highlighting the need for France to prepare for such eventualities by developing its own critical technologies.

Europe and Canada Draw Similar Lessons

The argument is not exactly new. Europe has been concerned for years about its dependence on the United States, whether technological or otherwise, and the European Union has increasingly emphasized the need to reduce the region's reliance on external suppliers in areas like chips, cloud computing, and AI. However, Anthropic's shutdown has made the situation more urgent, adding to an already deep unease regarding the reliability of the United States as an ally under Trump, from trade disputes to threats of NATO withdrawal. Attal stated that this issue will be central to the upcoming presidential elections in France, while members of the European Parliament have pointed out that the withdrawal of Mythos and Fable is proof that Europe must quickly realize its technological sovereignty.

Canada has drawn a similar lesson to that of Europe. Prime Minister Mark Carney stated that the situation highlights the risk of relying on a single partner for access to crucial resources like AI. "The situation we find ourselves in collectively right now with Mythos and Fable is something that can happen with excessive dependence on certain models," he said. "No one has done anything wrong in this situation. But we will have done something wrong if we accept this, do not learn the lesson, do not build, and do not diversify."

Diversification and Technological Independence

Other countries are already well advanced on this path. Beijing has long supported domestic AI companies, and China is one of the few places with models capable of credibly competing with products from American AI labs. However, in some areas, Chinese models lag behind their American counterparts, and Anthropic has accused Chinese competitors of using its models to train their own on an "industrial" scale. Part of the White House's decision to withdraw Mythos would stem from its belief that a group linked to China had gained access to the model.

Most governments and companies cannot compete with the scale and resources of leading labs in the United States or China. But sovereign AI does not always mean building the biggest or most powerful tools. France's Mistral models and Canada's Cohere demonstrate that strong efforts can come from outside these countries, even if the models cannot compete directly. Other countries, like Singapore and the United Arab Emirates, have focused on narrower but still strategic priorities, such as infrastructure or models that work better with local languages. Of course, there are also open-source models that could one day have capabilities similar to those of Mythos, which are difficult to control by a single party.

Trump may view the restriction of Mythos and Fable as a matter of national security. But the argument goes both ways, and with Washington now questioning whether AI is too important for everyone to access, other governments are wondering if they can afford to let Washington decide who has access.

Anthropic may soon bring Mythos and Fable back online. Restoring global trust in American AI is another matter. Regardless of how long the shutdown lasts, it has highlighted the fragility of access to American cutting-edge AI models. Many governments and companies did not like what they saw — and are determined to ensure that this does not happen again.