On March 20, Business Insider reported that Google had told employees that the company was convinced that cooperation with the U.S. Department of Defense was in line with the company’s AI principles and would be “more focused” on national security cooperation with governments. At an all-employee meeting held in January at DeepMind, Google’s AI company, leaders, including DeepMind CEO Demis Hassabis, responded to an employee’s question about how Google ensures that its collaboration with the Department of Defense and Boston Dynamics does not violate the company’s AI policies.
Google DeepMind CEO Hassabis
According to a recording of the meeting obtained by Business Insider, Tom Lue, Google DeepMind's vice president of global affairs, said that the company has established a "perfect process" to review specific usage scenarios and confirm whether partners have safeguards in place in terms of security, liability and privacy.
Tom Lu also reminded employees that the company had updated its AI principles in 2025, which deleted its previous commitment to "not use technology for the development of weapons or surveillance purposes."
"Our core criterion for this assessment is whether the benefits significantly outweigh the risks," he said.
The employee meeting was held earlier than Anthropic had an AI dispute with the US Department of Defense. Tom Lui said employees should expect more similar cooperation between the company and the government in the future.
"I also want to mention that this is an area that we will focus more on in the future. We are discussing with governments about their national security concerns." Tom Lu told employees. He said companies are having conversations about cybersecurity and biosecurity risks.
Google DeepMind CEO Hassabis once worried that Google would use DeepMind's technology for war. However, he told employees at the same conference that he was "very satisfied" with the balance Google has struck.
"Obviously, as we all see, the world is very complex, but I think we have a responsibility to work with democratically elected governments to provide our unique global leadership capabilities to help make the world safer and to bring benefits to the world." Hassabis said.
Google re-engaged with the Pentagon last year and won contracts for its AI and cloud products. Previously, Google abandoned a military cooperation in 2018 due to employee objections. This month, Google won another contract to deploy AI agents in the Department of Defense’s unclassified networks.
Asked for comment, a Google DeepMind spokesperson referred Business Insider to a blog post published last week. The article details Google's latest Pentagon contract. The blog post explains that Google tools are used for document drafting and project planning.