AI is dramatically increasing the speed and scale at which software vulnerabilities can be discovered. Anthropic recently reported that partners using Claude Mythos Preview found more than 10,000 high- or critical-severity vulnerabilities in a single month, while OpenAI is pursuing similar capabilities with GPT-5.6 Sol and its Daybreak program.
That’s great news for finding vulnerabilities, but it creates a new challenge for enterprise IT. Discovery can increasingly happen at machine speed, while remediation still requires teams to assess risk, source updates, package and test applications, manage exceptions, and safely deploy changes.
I recently explored this issue in articles published by The AI Journal and London Daily News, looking at what faster AI-assisted vulnerability discovery means for application management and how organizations can prepare. The answer isn’t simply to patch faster. Enterprises need better visibility into their application estate, repeatable paths for updating applications, automation for routine packaging and validation work, and ways to reduce exposure when an update can’t happen immediately.
The goal is to keep skilled people focused on the decisions that require judgment while automating the repeatable work around them. As AI makes vulnerability discovery increasingly scalable, that operational capacity is going to become much more important.
Read the articles:
- The AI Journal: When AI Finds Application Vulnerabilities Faster Than Enterprises Can Patch — July 17, 2026
- London Daily News: When AI Finds Vulnerabilities Faster Than Enterprises Can Patch Their Applications — August 14, 2026