Unverified50% confidenceOpinionExact time
The poisoning risks of AI programming can be understood from three levels: training data poisoning, context/prompt poisoning, and self-modifying loop poisoning.
1
Sources
50%
Confidence
Long-term
Relevance
9/18/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedAI systems have introduced entirely new dimensions of security risk, including erroneous decisions caused by model hallucinations, prompt injection attacks, training data poisoning, and AI agent privilege abuse.76% similarUnverifiedAI supply chain attack vectors include training data poisoning, backdoor implantation in pre-trained models, and tampering during model format conversion.76% similarUnverified研究人员已发现针对AI代码建议的提示注入、通过恶意仓库影响模型输出的数据投毒，以及利用Copilot上下文窗口实施的信息泄露攻击等新型攻击路径76% similarVerifiedAI代理操控计算机面临的安全风险包括提示注入攻击、权限升级和不可逆操作等问题72% similarUnverifiedEstablishing a regression testing mechanism is particularly critical in AI-assisted programming scenarios, because each code generation by the model may introduce hidden side effects, and it is used to constrain the non-deterministic behavior of the AI.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/931712API
curl https://kongchang.com/api/v1/knowledge/claims/931712MCP
get_claim(id=931712)