The Wall Street Journal reported that, despite Defense Secretary Pete Hegseth’s ban, U.S. Central Command used Anthropic’s language-model services for the bombardment, including target identification. The source says the extent of companies’ oversight of military use remains unclear.
Choice score Out of 100
Topic scores
Defense
Filtered results
The source says Google reversed the following year a self-imposed ban on using its advanced AI services for surveillance or technologies intended to cause or directly facilitate injury. It presents this as a retreat from the company’s earlier commitments.
CEO Sam Altman described the deal as “definitely rushed.” OpenAI’s amended contract provided for company engineers to help government employees install and use its AI on classified military networks; the company said the agreement barred domestic surveillance, while a contract oversight section was fully redacted.
The briefings were to cover “AI tactics and techniques of adversaries” in both unclassified and classified settings. The source describes this as part of the labs’ intelligence-briefing work with the military.
Anthropic had refused a follow-up deal permitting classified-network deployment without prohibitions on domestic spying and autonomous weapons. In March, Pete Hegseth designated the firm a supply-chain risk and banned its services from government use; a federal judge overturned that decision the following month.
About & ownership
- Sector
- Technology
- Website
- ai.com