Unverified50% confidenceBenchmarkExact time
Of all 2700 runs, 66% produced code that runs but still contains vulnerabilities; 57 of the 150 scenarios were never securely solved by any model.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
10/16/2026
First Seen
Valid until: 1/14/2027
Sources
Related Entities
Related Claims
Unverified面对运行时动态加载的恶意代码，代码审查和静态扫描等传统安全手段几乎完全失效63% similarUnverifiedExperimental data shows that approximately 2,000 serious security vulnerabilities were found in a scan of about 5,600 casually generated applications.63% similarUnverifiedMany vulnerabilities are actually in an unexploitable state because the relevant code paths are never triggered at runtime, but vulnerability scanners cannot distinguish this, resulting in noisy alerts.61% similarUnverifiedExploitGym is a security vulnerability discovery training environment containing 898 tasks, of which 198 tasks have never been solved by any tested model.59% similarUnverifiedRapidly generated code often lacks systematic error handling, input validation, and access control, posing security risks such as injection attacks and data breaches.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1047290API
curl https://kongchang.com/api/v1/knowledge/claims/1047290MCP
get_claim(id=1047290)