AI guardrails

Strict guardrails imposed by leading AI companies to prevent malicious use of their models are inadvertently hindering the work of legitimate offensive cybersecurity.