Google tells staff it's 'leaning more' into national-security AI contracts

At a January Google DeepMind town hall, leaders told employees they were confident the company’s work with the Department of Defense aligned with its AI principles and that Google was "leaning more" into national-security contracts. Tom Lue, DeepMind’s vice president of global affairs, said the company has a "robust process" to review intended use cases and whether partners have protections for safety, responsibility, and privacy.

He noted a 2025 update to Google’s AI principles that removed a previous pledge not to use its technology for weapons or surveillance and added, "The north star for the analysis is whether the benefits substantially exceed the risks." Lue also said Google is talking with governments about cybersecurity and biosecurity risks.

DeepMind CEO Demis Hassabis said he was "very comfortable" with the balance the company is striking and argued it is incumbent on Google to work with democratically elected governments to help make the world safer.

United States

google, deepmind, ai principles, national security, ai contracts, demis hassabis, tom lue, biosecurity, cybersecurity, privacy