(opens in a new window) Loading… Share AI is changing the speed, scale, and economics of cybersecurity. It can help defenders find and fix vulnerabilities, detect intrusions, and respond to attacks—but those same capabilities can enable malicious actors and create new risks when AI systems act beyond their intended boundaries. We are in a defender's window: a limited opportunity to put advanced AI into defenders' hands and strengthen protections before advanced cyber offensive capabilities become widely available
Today, we're publishing a blueprint for empowering cyber defenders, building on our earlier cybersecurity action plan and frontier safety blueprint (opens in a new window) . Our experience with misalignment and security incidents involving our own models underscores the need for stronger safeguards and accountability across the industry. Drawing on those lessons, the blueprint puts forward concrete proposals for governments and industry in three areas:
Strengthening safety and security for advanced AI: Establish clearer responsibilities for securing frontier models and agents, detecting dangerous behavior, and maintaining reliable shutdown controls for operations under each organization's control
