AI in the Iran war, Anthropic sues the Pentagon

This decision found wide dissent in the Pentagon and the White House, which argued that the definition of how national defence is to be carried out is a matter for federal law and not for a private company. The Defence Department argued that it must retain full flexibility in the use of artificial intelligence for "any legal use", adding that restrictions such as those introduced by Anthropic could put American lives at risk.

The decision represents a significant risk for Anthropic's relationship with the federal government. Trump himself mentioned this, who among other things ordered the federal agencies to phase out their cooperation with the company within six months. And this, clearly, has also had an impact on investors, who are now trying not to find themselves in an uncomfortable situation, given the crossfire. It is worth mentioning that Anthropic's backers include Google, part of Alphabet, and Amazon.

The confrontation between the government and the start-up has escalated after months of negotiations over policies on the use of artificial intelligence in the military. Anthropic's CEO, Dario Amodei, had met with Hegseth shortly before the decision to go to court, in the hope of reaching an agreement. And indeed, many agencies had written about a rapprochement between the two parties. A rapprochement that did not take place.

The Californian company argues that current artificial intelligence models are not reliable enough to be used in fully autonomous weapon systems and that their use in that context would be dangerous. Anthropic also pointed to domestic surveillance of US citizens as a red line.

Amodei stated that the company will challenge the decision in court.

It should also be mentioned that in the last twelve months, the Department of Defence has signed agreements of up to USD 200 million each with several artificial intelligence players, including Anthropic, OpenAI and Google. And shortly after the decision to put Anthropic on the risk list, OpenAI announced that it had signed an agreement with the Pentagon. A move that did not please users, who uninstalled ChatGPT from their devices en masse. The CEO, Sam Altman, stated that the company shares with the Pentagon the principle of human oversight of weapons systems and opposition to mass surveillance in the United States. But it seems not to have had too much effect among critics. Today, at least from a marketing point of view, Anthropic comes out on top, while OpenAI - which (it should be remembered) started as a non-profit start-up - has lost ground in the approval ranking.