Google is now more open to using AI for security and defense

Google

For years, Google made it clear—it didn’t want AI anywhere near weapons or surveillance. Back in 2018, the company even set strict rules against military use. This came after its employees protested against Project Maven, a program using AI for drone footage analysis. After facing severe backlash, Google backed off. It also promised not to use AI for millitary purposes. However, things seem to be changing. Google in its blog has announced updated AI policies.

In the blog titled Responsible AI Progress Report, Google quietly updated its policies. Now, it talks about working with governments and organizations to “protect people” and “support national security.” On the surface, it sounds like Google is just focusing on public safety. But in reality, when tech companies talk about 'national security,' it often includes things like surveillance and military defense. So, the company that once took a strong stand against military AI now seems to be leaving the door open.

Google argues that AI has become too important to ignore in areas like security and defense. It also believes that democracies—rather than authoritarian regimes—should lead AI development. That means working with like-minded governments to ensure AI is used “responsibly.” The problem? What’s considered “responsible” depends on who’s in charge.

To make sure its AI doesn’t go rogue, Google has introduced a Frontier Safety Framework. This includes stricter security measures and rules to prevent misuse. The framework is already in action for Google DeepMind’s latest AI, Gemini 2.0.

Not everyone is convinced by Google's explanation. Critics worry that this change could lead to AI being used for surveillance or even autonomous weapons. Google hasn’t said outright that it’s building military AI, but its updated policies suggest it’s no longer completely against the idea.