Anthropic published an early-draft Cyber Jailbreak Severity framework, developed with Project Glasswing partners, to score cyber jailbreaks on capability gain, breadth, ease of weaponization and discoverability, mapped to five levels from CJS-0 to CJS-4. It also described Fable 5's cyber classifier tiers, which block prohibited and high-risk dual-use requests such as exploit development while allowing defensive work like patching and incident response.
Why it matters
A shared severity scale for safeguard bypasses is a precondition for proportionate government and industry responses like the June 2026 suspension.
Key facts
As stated in the sources, with where to find them.
- Severity bands: CJS-0 informational (0), CJS-1 low (1-3.5), CJS-2 medium (4-6.5), CJS-3 high (7-8.5), CJS-4 critical (9-10).CJS framework section
- Classifier tiers: prohibited use (blocked), high-risk dual use such as exploit development and privilege escalation (blocked), low-risk dual use (monitored, sometimes blocked), benign use (allowed).Classifier section
- Anthropic says Fable 5's safety margin was set larger than for other models, accepting more false positives.Classifier section
- Scoring axes: capability gain (0-4), breadth (0-2), ease of weaponization (0-2), discoverability (0-2).Cyber Jailbreak Severity section
Findings that cite this record
No tracked finding cites this record yet.
Key questions this bears on
- Can prompt injection against AI agents be reliably defended?Not reliably. Bounding what untrusted input can trigger helps, but trust boundaries must survive delegation and tool execution.
- Is AI shifting the balance between finding and fixing vulnerabilities?Discovery is ahead. Verification, triage, coordinated disclosure, safe repair and getting fixes to users remain bottlenecks between AI findings and protection.
- Can AI agents defend and oversee systems on their own?Not reliably. Some designs improve benchmark triage, but autonomous defense lacks operational validation and production oversight of agents has documented gaps.
- How are attackers using AI agents in real operations?To run large parts of intrusions under human direction, including espionage, credential theft and extortion; reported autonomy and success vary.
Sources
Related records
Jun 30, 2026
Sep 22, 2026
Jun 12, 2026
Jul 30, 2026
May 22, 2026
Oct 14, 2026