Google is negotiating with the Pentagon to integrate its Gemini AI system into defense operations, amid employee opposition and ethical concerns.
Google is currently in negotiations with the U.S. Department of Defense to incorporate its Gemini AI system into Pentagon operations, covering all levels of information classification, from public to highly sensitive data. This potential agreement would enable the Pentagon to leverage Google's technology for various lawful applications.
As part of the discussions, Google has proposed measures to ensure that Gemini will not be employed for mass surveillance or the creation of autonomous weaponry. A spokesperson from the Pentagon refrained from commenting specifically on the negotiations but emphasized the department's commitment to collaborating with private sector firms to enhance AI technology across all classification levels.
In a related development, Google previously signed a contract with the Pentagon on April 28, 2026, allowing the use of AI for lawful government functions. This agreement obliges Google to assist in setting up security parameters and AI filters as requested by the government, but it explicitly states that the AI system should not be used for mass surveillance or autonomous weapons without appropriate human oversight. However, it does not provide the Pentagon with any veto power over lawful operational decisions.
Despite internal resistance, with over 600 employees urging CEO Sundar Pichai to stop Pentagon AI use in classified settings, Google is moving forward with the deal. This internal dissent echoes the company's past withdrawal from Project Maven in 2018, a program aimed at using AI for drone footage analysis, which was scrapped after significant employee pushback.
In contrast to Google, the Pentagon has faced challenges with other AI firms, such as Anthropic, which recently encountered legal disputes over ethical considerations concerning its AI technologies. The Department of Defense has been keen on integrating AI solutions from various contractors, including Anthropic, Google, OpenAI, and xAI, with contracts amounting to $200 million awarded for security AI development.
As negotiations with Google continue, the Pentagon aims to accelerate its decision-making processes by integrating AI into its operational framework. The military seeks to utilize AI's capabilities for tasks like operational planning and targeting, despite the inherent risks of errors and misinformation associated with neural networks in combat scenarios.
