Vendor-reported results for one lab's safeguards, built for chemical-weapons and other CBRN content rather than cyber misuse. Anthropic's September 2026 Opus 5.5 card adds cyber-specific safeguards and red-team results, also vendor-reported.
Reported: Stated by one source and not yet corroborated or challenged.
Evidence
Automated jailbreak success on Claude 3.5 Sonnet fell from 86% to 4.4%; in the Feb 3-10, 2025 public demo one participant found a universal jailbreak.
Over 1,700 red-teaming hours and 198,000 attempts found no universal jailbreak; about 1% compute overhead.
Anthropic reports cyber-specific safeguards for Opus 5.5 (an activation probe plus two classifiers) and no evidence of a critical-severity jailbreak from internal and contracted red teaming; a contractor, Trajectory Labs, reported 13 candidate breaks and, on one task, a privilege-escalation chain built across over 100 separate contexts, but no universal jailbreak. Vendor-reported.
Status history
- 2025-02-03ReportedAnthropic Constitutional Classifiers. · record