Google Expands Pentagon AI Access Following Anthropic's Refusal

Recent industry reports confirm that Google has officially granted the U.S. Department of Defense (DoD) access to its artificial intelligence systems on classified networks, effectively greenlighting all lawful military applications.

This agreement follows a highly publicized standoff between Anthropic and the Trump administration, stemming from the AI developer’s refusal to offer the Pentagon unrestricted access. While the DoD seeks broad deployment capabilities, Anthropic insists on implementing strict safety guardrails to prevent its models from being utilized for domestic mass surveillance or autonomous weapon systems.

Due to Anthropic’s firm stance on these use cases, the Pentagon designated the company as a “supply chain risk,” a classification historically reserved for foreign adversaries. The two parties are now embroiled in ongoing litigation, with a federal judge recently issuing a preliminary injunction to temporarily block the designation as legal proceedings continue.

Capitalizing on the situation, Google has emerged as the third major AI firm to secure a defense contract after OpenAI and xAI. The newly signed agreement includes provisions stating that Google does not intend to deploy its AI for domestic surveillance or lethal autonomous weapons, mirroring language found in OpenAI’s contract. However, as highlighted by The Wall Street Journal, it remains unclear whether these contractual safeguards carry legally binding weight or can be effectively enforced.