The number of reported security vulnerabilities has skyrocketed since AI models started searching for bugs, with a record 1,500 high-severity vulnerabilities reported in June 2026, more than triple the previous monthly record. This surge is largely due to AI-powered tools like Anthropic’s Claude Mythos Preview and OpenAI’s Daybreak program, which can automatically detect software vulnerabilities. As a result, organizations are now able to identify and fix critical bugs at an unprecedented rate. Why it matters: this trend signals a significant shift in the cybersecurity landscape, where AI-driven discoveries are revolutionizing the way we detect and address security threats.