Google is deploying Gemini AI agents to the Department of Defense's more than 3 million civilian and military employees, initially on unclassified networks. The rollout follows Google's recent shift in AI principles regarding weapons and surveillance applications. Emil Michael, Under Secretary of Defense for Research and Engineering, confirmed talks are underway to expand access to classified and top-secret systems.

Eight pre-built agents will automate tasks like summarizing meeting notes, building budgets, and checking proposed actions against the national defense strategy. Jim Kelly, Google Vice President, announced that Defense Department personnel can create custom agents using natural language through the Pentagon's GenAI.mil portal. Since December, 1.2 million Defense Department employees have used Google's AI chatbot for unclassified work, running 40 million unique prompts and uploading over 4 million documents.

Despite this rapid adoption, training has lagged significantly—only 26,000 people have completed AI training since December, though future sessions are fully booked. This expansion comes as the Pentagon broadens its AI partnerships following its standoff with Anthropic, which refused to remove guardrails against domestic surveillance and autonomous weapons. Google itself faced internal backlash in 2018 over Project Maven, a drone video analysis program, but has since loosened restrictions on military work while other tech companies like OpenAI and xAI also strike defense deals.