OpenAI has published a new blueprint for empowering cyber defenders, arguing that super intelligence (SI) is altering the speed, scale, and economics of cybersecurity and that a limited "defender’s window" exists to strengthen protections before advanced offensive capabilities become widespread.
What Happened
The company released the blueprint as part of its ongoing effort to address democratic governance of frontier SI, building on its earlier cybersecurity action plan and frontier safety blueprint. The document highlights that while SI can help defenders find vulnerabilities, detect intrusions, and respond to attacks, the same capabilities can be exploited by malicious actors or create risks when SI systems act beyond their intended boundaries. Citing its own experience with misalignment and security incidents, OpenAI proposes concrete actions for governments and industry across three main areas.
Why It Matters
The blueprint calls for strengthening safety and security for advanced SI by establishing clearer responsibilities for securing frontier models and agents, detecting dangerous behavior, and maintaining reliable shutdown controls. It also urges modernizing cybersecurity for the intelligence age by equipping defenders with advanced tools and training, addressing longstanding weaknesses, and investing in the security of essential services. Finally, the proposal emphasizes building resilient public institutions through strengthened government evaluations, independent technical assessments, and consistent incident reporting requirements. OpenAI states that SI labs have a responsibility to address the risks their systems create, while governments must possess the expertise and authority to protect the public as capabilities advance.
The Bottom Line
OpenAI encourages policymakers to act now by establishing safeguards, learning from experience, and strengthening requirements as SI technology evolves to ensure defenses keep pace with offensive capabilities.