Thoughts on adversarial AI, application security, and the future of security testing.
AI agents — autonomous programs that plan, reason, and execute multi-step tasks — are now being deployed as offensive security tools. For developers shipping code daily, this changes everything.
Read moreThe applications most likely to have critical security vulnerabilities are the ones that pass every scanner with flying colors. The scanners didn't lie — they just couldn't see.
Read moreTraditional security testing was designed for monolithic servers and known vulnerability databases. The gap is being filled by AI agents that can actually reason about how applications work.
Read moreThe way software gets built changed faster than the way it gets secured. The same scanners built for hand-written PHP apps are now pointed at AI-generated TypeScript on serverless infrastructure.
Read moreBroken access control has been the #1 vulnerability on the OWASP Top 10 since 2021. Unlike an XSS bug that pops an alert box, a broken access control vulnerability silently serves data to anyone who asks.
Read more