3 months ago
1
The Pentagon is considering ending its narration with artificial intelligence institution Anthropic implicit its insistence on keeping immoderate restrictions connected however the U.S. subject uses its models, Axios reported connected Saturday (February 14, 2026), citing an medication official.
The Pentagon is pushing 4 AI companies to fto the subject use their tools for “all lawful purposes,” including in areas of weapons development, quality postulation and battlefield operations, but Anthropic has not agreed to those presumption and the Pentagon is getting fed up aft months of negotiations, according to the Axios report.
The different companies included OpenAI, Google and xAI.
An Anthropic spokesperson said the institution had not discussed the usage of its AI exemplary Claude for specific operations with the Pentagon. The spokesperson said conversations with the U.S. authorities so acold had focused connected a circumstantial acceptable of usage argumentation questions, including hard limits around afloat autonomous weapons and wide home surveillance, nary of which related to existent operations.
The Pentagon did not instantly respond to Reuters’ petition for comment.
Anthropic’s AI exemplary Claude was utilized successful the U.S. military’s cognition to seizure former Venezuelan President Nicolas Maduro, with Claude deployed via Anthropic’s concern with information steadfast Palantir, the Wall Street Journal reported connected Friday (February 13, 2026).
Reuters reported on Wednesday (February 11, 2026) that the Pentagon was pushing apical AI companies including OpenAI and Anthropic to marque their artificial quality tools disposable connected classified networks without galore of the modular restrictions that the companies use to users.
