TLDR Tech / InfoSec
Flaw-Finding AI Assistants Face Criticism for Speed, Accuracy
Anthropic's Claude Code Security and OpenAI's Aardvark raise concerns about being slow, expensive, and noisy compared to established SAST tools, especially when the same AI both writes and reviews code. Experts argue these assistants should augment, not replace, existing pipelines, as AI-driven “vibe coding” accelerates insecure code and growing security debt. The real value lies in AI-enhanced remediation workflows and the potential revival of interactive, AI-powered code review.
Read the full story