Pentagon Shifts AI Strategy Toward Google Gemini Amid Anthropic Blacklisting

The U.S. Department of Defense (DOD) is intensifying its partnership with Google, integrating the Gemini AI model into classified projects to enhance wartime capabilities. Pentagon AI chief Cameron Stanley confirmed the expansion, emphasizing that the adoption of Gemini is already saving thousands of man-hours on a weekly basis. This strategic pivot comes as the DOD distances itself from Anthropic, which has been designated as a supply chain risk. While a San Francisco judge previously blocked a ban on Anthropic's Claude model for other government agencies, a Washington D.C. appeals court recently denied a request to stop the DOD's specific blacklisting of the firm, leaving Anthropic excluded from defense contracts. Stanley highlighted a broader commitment to a multi-vendor ecosystem, noting that the Pentagon is also collaborating with OpenAI and other providers. He cautioned against overreliance on a single software vendor, stating that diversification is critical for operational resilience and avoiding systemic vulnerabilities. However, the partnership faces internal headwinds at Google. More than 700 employees have signed a letter to CEO Sundar Pichai urging the company to reject classified workloads, citing concerns that the technology could be used in harmful or inhumane ways. Looking forward, the DOD remains on high alert following the rollout of Anthropic's Mythos model. Stanley described the release as a 'wakeup call' due to its advanced cyber capabilities, signaling that the Pentagon must accelerate its own AI-enabled capabilities to match evolving global threats.