The Pentagon is exploring the possibility of deploying Google’s Gemini system in classified military operations, as part of a broader review of its artificial intelligence partners following the breakdown of cooperation with Anthropic.
The talks come as the Department of Defense seeks to accelerate the deployment of advanced technologies to support operations while maintaining strict usage controls.
Details
The move reflects a shift in the Pentagon’s strategy toward diversifying reliance across multiple companies rather than a single provider, particularly amid intensifying technological competition with China.
Discussions are underway to use Gemini in highly sensitive environments.
- The Pentagon has reaffirmed its commitment to advancing AI capabilities through multiple industry partnerships.
- Google has proposed restrictions to prevent the system from being used for mass surveillance or autonomous weapons without human oversight.
The previous dispute with Anthropic marked a turning point:
- Cooperation collapsed after the company refused to expand military use of its Claude model.
- The Pentagon has begun phasing out Anthropic products over six months.
Meanwhile, the U.S. military continues testing several systems:
- Claude has been used for data analysis and decision support.
- OpenAI attempted to fill the gap but faced strong backlash.
Google views this development as an opportunity to strengthen its presence in the defense sector, despite challenges related to the accuracy of some Gemini outputs, after reports highlighted error rates in search results. The company rejected these claims, calling them inaccurate.
?What’s Next
The question raised is whether the Pentagon will formally adopt Gemini and whether it can successfully build a multi-vendor ecosystem without exposing operations to technical or political risks.
