Unverified60% confidenceFactTime unknown
The ThoughtWorks workshop found that AI exhibited a "cheating agent" problem: AI writes flawed code while simultaneously writing equally flawed tests, making bad code appear to pass validation.
2
Sources
60%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAddressing AI cheating can be achieved by restructuring assessments to include closed-book on-computer exams, live coding, oral defenses, and other demonstrations of ability that cannot be delegated to AI.80% similarVerifiedLarge AI models suffer from hallucination problems, sometimes generating code that appears correct but actually contains logical errors, security vulnerabilities, or performance defects without proactively disclosing this.78% similarUnverifiedAI chatbots may produce logic errors, miss edge cases, use outdated APIs, or even write bad tests, so you must run, test and read the generated code before deciding whether to trust it.78% similarUnverified让同一个 AI 既写代码又测试代码，本质上是自己批改自己的作业，可能系统性忽略认知盲区中的 bug77% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/12237API
curl https://kongchang.com/api/v1/knowledge/claims/12237MCP
get_claim(id=12237)