Tag
StealthBench measures operational stealth in autonomous offensive-security agents across six OPSEC dimensions, using a 3-model LLM judge panel. Results show no model exceeds 54% safe success rate, indicating systematic OPSEC failures.
A third party breached The Intercept's Signal tip line and impersonated the publication to solicit whistleblowers, putting sources at risk. The Intercept has since switched to a new Signal account but has not disclosed full details of the breach.