[KongchangAI]
Unverified50% confidenceFactExact time

A study in March 2026 comparing tens of thousands of AI-generated and human-written code units in real codebases found that AI code triggered slightly fewer scan warnings and also contained fewer hardcoded secrets.

1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen

Sources

Vibe Coding Security Checklist: 12 Essential Checks

Related Entities

Mao et al.person

Related Claims

UnverifiedData leakage almost never causes an error; the code runs normally and the metrics look great, making it one of the hardest-to-automatically-detect pitfalls in AI-generated code.76% similarUnverifiedGitHub的研究发现AI辅助编写的代码中安全缺陷引入率与手写代码相近，但因开发者对AI产出存在自动化偏见，这些缺陷更难在审查阶段被发现75% similar

Cite This Claim

Stable URI
https://kongchang.com/claim/1002917
API
curl https://kongchang.com/api/v1/knowledge/claims/1002917
MCP
get_claim(id=1002917)