Google's AI for Cyber-Warfare Restricted to Governments

Google has created a special AI called Gemini 3.5 Flash Cyber that can find security problems in computer code, create proofs to show these problems are real, and even fix them.

This AI is very powerful and can do its work in just a few hours.

Google has decided that only governments and trusted partners can use this AI because it can also create ways to exploit these security problems.

This means that the AI can be used both to protect and to attack, so Google wants to make sure it is used responsibly.

The AI was released along with two other general-purpose models, but this one is special because it is designed for cybersecurity work.

Google believes that by restricting access, they can help protect important systems and prevent bad actors from using the AI for harmful purposes.