ResearchTechCrunch

How AI guardrails are impeding the work of offensive cybersecurity researchers

#cybersecurity#ai#research#vulnerabilities#guardrails

English

AI guardrails implemented by major companies to prevent misuse by malicious hackers are now hindering the work of offensive cybersecurity researchers. These restrictions limit researchers' ability to test and exploit vulnerabilities, leading some to revert to open-source models without guardrails.

中文

大型公司实施的AI护栏旨在防止恶意黑客滥用，但现在却妨碍了进攻性网络安全研究人员的工作。这些限制限制了研究人员测试和利用漏洞的能力，导致一些人转向没有护栏的开源模型。