Anthropic CEO Dario Amodei is pursuing a last-minute compromise with the Pentagon to prevent his AI company from being blacklisted by the US government. Negotiations are underway to govern military access to Anthropic's AI models.
A resolution would allow the Pentagon to continue using Anthropic's technology and avert a designation that would sever defense contracts. The dispute arose after reports indicated Anthropic's AI, Claude, was used in a US operation in Venezuela, raising concerns about compliance with the company's acceptable use policy.
Anthropic's opposition to autonomous weapons and mass surveillance clashes with the Pentagon's desire for broader permission for "lawful" AI use. Following CEO Amodei's rejection of an ultimatum, federal agencies were ordered to halt Anthropic's technology use, and the firm was designated a national security supply chain risk.
Amodei has accused the Pentagon and a rival AI firm of misrepresenting the situation. Anthropic, a competitor in a $200 million Pentagon deal for agentic AI, risks a significant setback if this military foothold is lost.