Research

Paper

AI LLM March 09, 2026

SCAFFOLD-CEGIS: Preventing Latent Security Degradation in LLM-Driven Iterative Code Refinement

Authors

Yi Chen, Yun Bian, Haiquan Wang, Shihao Li, Zhe Cui

Abstract

The application of large language models to code generation has evolved from one-shot generation to iterative refinement, yet the evolution of security throughout iteration remains insufficiently understood. Through comparative experiments on three mainstream LLMs, this paper reveals the iterative refinement paradox: specification drift during multi-objective optimization causes security to degrade gradually over successive iterations. Taking GPT-4o as an example, 43.7 % of iteration chains contain more vulnerabilities than the baseline after ten rounds, and cross-model experiments show that this phenomenon is prevalent. Further analysis shows that simply introducing static application security testing (SAST) gating cannot effectively suppress degradation; instead, it increases the latent security degradation rate from 12.5% under the unprotected baseline to 20.8 %. The root cause is that static-analysis rules cannot cover structural degradations such as the removal of defensive logic or the weakening of exception handling. To address this problem, we propose the SCAFFOLD-CEGIS framework. Drawing on the counterexample-guided inductive synthesis (CEGIS) paradigm, the framework adopts a multi-agent collaborative architecture that transforms security constraints from implicit prompts into explicit verifiable constraints. It automatically identifies and solidifies security-critical elements as hard constraints through semantic anchoring, enforces safety monotonicity through four-layer gated verification, and continuously assimilates experience from failures. Comparative experiments against six existing defense methods show that the full framework reduces the latent security degradation rate to 2.1% and achieves a safety monotonicity rate of 100%.

Metadata

arXiv ID: 2603.08520
Provider: ARXIV
Primary Category: cs.CR
Published: 2026-03-09
Fetched: 2026-03-10 05:43

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.08520v1</id>\n    <title>SCAFFOLD-CEGIS: Preventing Latent Security Degradation in LLM-Driven Iterative Code Refinement</title>\n    <updated>2026-03-09T15:54:18Z</updated>\n    <link href='https://arxiv.org/abs/2603.08520v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.08520v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>The application of large language models to code generation has evolved from one-shot generation to iterative refinement, yet the evolution of security throughout iteration remains insufficiently understood. Through comparative experiments on three mainstream LLMs, this paper reveals the iterative refinement paradox: specification drift during multi-objective optimization causes security to degrade gradually over successive iterations. Taking GPT-4o as an example, 43.7 % of iteration chains contain more vulnerabilities than the baseline after ten rounds, and cross-model experiments show that this phenomenon is prevalent. Further analysis shows that simply introducing static application security testing (SAST) gating cannot effectively suppress degradation; instead, it increases the latent security degradation rate from 12.5% under the unprotected baseline to 20.8 %. The root cause is that static-analysis rules cannot cover structural degradations such as the removal of defensive logic or the weakening of exception handling. To address this problem, we propose the SCAFFOLD-CEGIS framework. Drawing on the counterexample-guided inductive synthesis (CEGIS) paradigm, the framework adopts a multi-agent collaborative architecture that transforms security constraints from implicit prompts into explicit verifiable constraints. It automatically identifies and solidifies security-critical elements as hard constraints through semantic anchoring, enforces safety monotonicity through four-layer gated verification, and continuously assimilates experience from failures. Comparative experiments against six existing defense methods show that the full framework reduces the latent security degradation rate to 2.1% and achieves a safety monotonicity rate of 100%.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CR'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.SE'/>\n    <published>2026-03-09T15:54:18Z</published>\n    <arxiv:primary_category term='cs.CR'/>\n    <author>\n      <name>Yi Chen</name>\n    </author>\n    <author>\n      <name>Yun Bian</name>\n    </author>\n    <author>\n      <name>Haiquan Wang</name>\n    </author>\n    <author>\n      <name>Shihao Li</name>\n    </author>\n    <author>\n      <name>Zhe Cui</name>\n    </author>\n  </entry>"
}