Brian Clark works through the types of threat outlined by OWASP and beyond, with concrete steps for protecting AI projects from the biggest ones.
It includes live-hacking AI applications and LLMs, which tends to be more persuasive about the gravity of AI security than a slide about it.
Snyk’s Brian Clark has contributed to the Web Almanac security chapter and spends his time on how developers actually secure applications, so the framing is practical rather than theoretical — what to do on Monday, not a taxonomy.
