{"slug": "anthropic-wants-to-grade-ai-jailbreaks-like-cves-here-s-the-framework", "title": "Anthropic wants to grade AI jailbreaks like CVEs. Here's the framework.", "summary": "Anthropic has published a structured framework for rating the severity of AI jailbreaks, called the Cyber Jailbreak Severity (CJS) scale, which ranges from CJS-0 (Informational) to CJS-4 (Critical). The company aims to standardize communication about jailbreak risks, similar to how CVSS scores work for software vulnerabilities. Anthropic has also launched a HackerOne program for researchers to submit jailbreaks for review.", "body_md": "Anthropic has re-deployed Claude Fable 5 and used the moment to publish something the industry has been missing: a structured framework for talking about how dangerous an AI jailbreak actually is.\n\nThink CVE severity scores, but for AI. The Cyber Jailbreak Severity (CJS) scale runs from CJS-0 (Informational — no real uplift) to CJS-4 (Critical — domain-expert-level output that meaningfully accelerates real attacks). Anthropic is calling it an early draft and asking for feedback, but the intent is clear: standardize the language so AI developers and governments can actually communicate about these risks.\n\n\"There is no agreed-upon framework for describing a given jailbreak's severity. Such a framework would allow AI developers to speak to governments (and vice versa) in consistent terms about the risks posed by each jailbreak.\"\n\nAnthropic has laid out a four-tier classifier system for cybersecurity use cases:\n\nOne nuance worth flagging: Fable 5's safety margin is deliberately larger than previous models. That means more false positives — legitimate requests getting blocked — but Anthropic is prioritising caution at launch.\n\nThe CJS scale grades jailbreaks on four axes:\n\nThe bands are exponential, not linear — each step is several times more serious than the last. CJS-4 is reserved for jailbreaks that produce domain-expert-level outputs that aren't otherwise obtainable and require little expertise to weaponize.\n\nThey've also launched a [HackerOne program](https://hackerone.com/anthropic-cyber-jailbreak/) where researchers can submit Fable 5 jailbreaks for review.\n\nAnthropic is trying to do for AI jailbreak severity what CVSS did for software vulnerabilities — create a shared vocabulary that makes it possible to triage, prioritize, and communicate risk consistently.\n\nThat has real implications. If this framework (or something like it) gets adopted, it becomes the language that regulators, procurement teams, and incident responders use when AI systems are involved in a breach. It changes how liability gets discussed.\n\nThe fact that they're publishing it openly and asking for feedback at [cyber-safeguards@anthropic.com](mailto:cyber-safeguards@anthropic.com) suggests this is a genuine standards-building effort, not just a PR move.\n\n*Source: Anthropic — More details on Fable 5's cyber safeguards and our jailbreak framework*\n\n*✏️ Drafted with KewBot (AI), edited and approved by Drew.*", "url": "https://wpnews.pro/news/anthropic-wants-to-grade-ai-jailbreaks-like-cves-here-s-the-framework", "canonical_source": "https://dev.to/thegatewayguy/anthropic-wants-to-grade-ai-jailbreaks-like-cves-heres-the-framework-5fk6", "published_at": "2026-07-10 13:06:59+00:00", "updated_at": "2026-07-10 13:13:40.218884+00:00", "lang": "en", "topics": ["ai-safety", "ai-policy", "large-language-models", "ai-research"], "entities": ["Anthropic", "Claude Fable 5", "HackerOne", "CJS scale", "Drew"], "alternates": {"html": "https://wpnews.pro/news/anthropic-wants-to-grade-ai-jailbreaks-like-cves-here-s-the-framework", "markdown": "https://wpnews.pro/news/anthropic-wants-to-grade-ai-jailbreaks-like-cves-here-s-the-framework.md", "text": "https://wpnews.pro/news/anthropic-wants-to-grade-ai-jailbreaks-like-cves-here-s-the-framework.txt", "jsonld": "https://wpnews.pro/news/anthropic-wants-to-grade-ai-jailbreaks-like-cves-here-s-the-framework.jsonld"}}