AI Finds More Vulnerabilities, but Remediation Is Falling Behind

AI is helping security teams discover vulnerabilities faster, but TechRadar reports that validation, prioritization and remediation are not keeping pace.

Over the past year, teams on the platform cut the time needed to fix a critical vulnerability by roughly 50%. During the same period, their backlog of unresolved critical vulnerabilities grew nearly 29-fold. The figures can coexist when discovery accelerates faster than remediation and engineering capacity.

AI models can read code and search thousands of assets for familiar vulnerability patterns. However, each finding still needs confirmation, severity assessment and review in its operating context. It must then reach the right engineering team, compete with existing priorities, be fixed and be retested.

TechRadar describes Continuous Threat Exposure Management (CTEM) as a process for understanding the attack surface, identifying weaknesses, verifying which are exploitable and directing remediation toward the greatest business risk.

Human judgment remains important because businesses know which services generate revenue, where regulated data resides and which controls are already in place. Researchers also help identify attack paths formed by several weaknesses. Researchers earned more than $47 million on the H1 Platform in the first half of this year, up more than 25% year on year, according to the article. TechRadar notes that rising aggregate earnings do not mean every researcher benefits equally as AI absorbs routine findings.