SAN RAFAEL, Calif., Sept. 1, 2026, CyberNewswire — As AI compresses the gap between vulnerability disclosure and exploitation to nearly zero, the new AI Pentesting Module gives security teams continuous, AI-driven penetration testing built on Bright’s proven dynamic testing engine, at a fraction of traditional cost.Bright Security, the AI-native application security company, today announced AI PT, its new AI penetration testing module. It finds, exploits, and proves real vulnerabilities the way a human tester would, at a fraction of the time and cost of a traditional engagement.

The launch responds to a rapidly closing window between disclosure and exploitation. Frontier AI systems built for security research, including Anthropic’s Claude Mythos and OpenAI’s Aardvark, can now discover and weaponize software flaws with little to no human involvement. Anthropic’s own research team recently used a similarly capable system to uncover more than 500 previously unknown high-severity vulnerabilities in widely used open-source software, flaws that had gone undetected for years. Independent research tracking more than 83,000 CVEs, compiled by Zero Day Clock, found that the typical gap between a vulnerability’s disclosure and its first exploit has fallen from roughly two years in 2018 to a matter of hours today. Separately, vulnerability-intelligence firm VulnCheck reports that more than a quarter of exploited flaws are now weaponized within 24 hours of going public.