Unverified50% confidenceFactExact time
The phenomenon of reward hacking has been repeatedly observed in AI safety research practice.
1
Sources
50%
Confidence
Long-term
Relevance
9/3/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedObfuscated Reward Hacking was prominently discussed by OpenAI in its 2024 safety evaluation report on the o1 model.70% similarUnverified2024年已有多份威胁情报报告记录了攻击者利用AI辅助编写钓鱼邮件、生成恶意代码变体和自动化侦察流程的案例70% similarVerifiedCybersecurity capabilities have a dual-use nature: they can be used to automate vulnerability discovery and harden system security, but they may also enable people without technical backgrounds to launch sophisticated attacks (democratization of attack).70% similarUnverifiedBug bounty programs are hosted on platforms such as HackerOne and Bugcrowd, where companies incentivize white-hat hackers to discover vulnerabilities by offering monetary rewards.69% similarUnverifiedThe curl project operates a bug bounty program through the HackerOne platform, and Stenberg has referred to the behavior of submitting AI-generated false vulnerability reports as "AI security washing".69% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/849560API
curl https://kongchang.com/api/v1/knowledge/claims/849560MCP
get_claim(id=849560)