IBM on Monday, Oct. 19, 2026, launched X-Force Advanced Proactive Security, a security testing service that applies frontier AI models to penetration testing while keeping IBM’s X-Force testers in charge of the work. The company said the service uses frontier AI “via OpenAI’s Daybreak program and other best-of-breed solutions,” according to an IBM post by Ray Sims of X-Force Offering Strategy.

The offering combines expert-led tests with autonomous, AI-driven testing for lower-risk work, all run on a single platform. IBM said the aim is to give exposure management programs continuous, validated evidence of what an attacker could do, which it said those programs often lack.

The problem IBM describes

The post argues that AI-driven tools can produce findings faster than security teams can validate them, which adds to workloads when the tools run without control. And penetration testing is still scoped and scheduled around a snapshot in time.…