How AI guardrails are impeding the work of offensive cybersecurity researchers

AI safety guardrails built into systems like ChatGPT are making it harder for cybersecurity researchers to find and fix vulnerabilities responsibly. This matters to IT professionals and the security of systems everyone relies on.