Anthropic challenges US government restrictions on AI technology use

The US AI company Anthropic attempted to challenge a decision by the US Department of Defense to ban the use of its technology in government institutions. The company had been given 30 days to file an appeal after the Pentagon classified its products as a ‘supply chain risk’.

However, instead of following the proper procedures and sending the appeal to the designated official address, the head of Anthropic’s legal team emailed it directly to two DoW officials.

Reports cited by multiple outlets suggested disagreements between Anthropic and US defence authorities over potential military applications of AI systems and related safety restrictions. The reports referred to discussions about military and intelligence-related AI applications. After Pentagon asked Anthropic to bend its rules and company refused, Anthropic was latter classified as a ‘supply chain risk’.

Reports described the designation as an unusual step involving a major US AI company. According to reports, federal agencies were instructed to suspend use of Anthropic technologies following the classification decision.

Anthropic CEO Dario Amodei has said the company previously cooperated with the US military in areas such as intelligence analysis, modelling and simulation, operational planning, and cyber operations. However, he has also argued that the government’s action was not legally justified and indicated that the company has no choice but to challenge it in court.

Amodei went on to say that Anthropic does not believe private companies should be involved in operational military decision-making, particularly when systems could enable fully autonomous weapons or mass surveillance.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!