IBM AI: Boosting Cyber Defenses Against Advanced Threats

The biggest risk for enterprises today isn't just finding vulnerabilities, but how quickly they can respond to cyberattacks. Frontier AI models are shrinking attack timelines from weeks to days or even hours. Most security teams, however, still rely on manual reviews. IBM is working to close this gap by participating in Project Glasswing, an industry effort to protect critical software from AI-driven threats. They are hardening their own products and sharing findings. Attackers are using AI to speed up cyberattacks, so security teams are beginning to deploy their own frontier models, like Claude's Mythos Preview, to identify vulnerabilities. IBM uses a multi-model approach, integrating AI models, including frontier models, into their defenses. They are weaving AI into their core security principles to keep pace with evolving threats. This collaboration and technological advancement are crucial for strengthening collective defenses.