Public story · 2026-07-02 · high
Anthropic, Google, Microsoft, Amazon draft jailbreak severity score
The framework mirrors CVSS's approach to software bugs, letting buyers compare AI jailbreak failures across vendors.
Why now: The four labs are floating this proposal before any of them has said who actually sets the score.
Story
Anthropic proposed a shared severity scale for jailbreak failures, with Google, Microsoft, and Amazon signing on, according to the company's announcement.
Buyers evaluating AI vendors have no common yardstick for how bad a jailbreak actually is, so every disclosure gets judged on its own terms.
The framework mirrors CVSS, the scale that standardized vulnerability reporting for the software industry. A shared score would let enterprise teams triage AI jailbreaks the same way security teams already triage CVEs, instead of comparing incompatible incident writeups across vendors.
Yes, but the four labs writing the scale are the same four labs whose models get graded on it. The proposal doesn't say who holds scoring authority or whether each lab ends up scoring its own jailbreaks.
I'd rather compare failure modes across vendors on a common scale than take each lab's word for how bad its own jailbreak was. A severity scale graded by the model builders is a marketing tool, not a safety standard. It only counts once someone outside those four companies controls the rubric. Watch who ends up owning it.
Anthropic posted the proposal on its news page while jailbreak severity still has no shared vocabulary across the industry.
Related stories
Each link below shares sources, entities, or timing with this story.
Shared entities / Shared topic / Earlier coverage / Tension
Google, Microsoft, and xAI Give US Government Pre-Launch Model Access.
Both cover Anthropic, Google, Microsoft; overlapping topics (anthropic, model); earlier Anthropic coverage from 2026-05-10.
Shared entities / Earlier coverage
MCP Security: Three Critical Vulnerabilities in One Day
Both cover Anthropic, CVSS, Google, Microsoft; earlier Anthropic coverage from 2026-02-28.
Shared entities / Earlier coverage / Tension
The Linux Foundation launched Akrites to defend open source against AI-enabled threats.
Both cover Anthropic, Google, Microsoft; earlier Anthropic coverage from 2026-06-27; pushes against this story (against).
OpenAI launches a $150M partner network to win enterprise agent rollouts.
Both cover Anthropic, Google, Microsoft; earlier Anthropic coverage from 2026-06-15; pushes against this story (against).
Shared entities / Same source / Earlier coverage
Apple opened its Foundation Models framework to Claude and Gemini, and quietly shipped it on Linux
Both cover Anthropic, Google; cite the same source (Anthropic); earlier Anthropic coverage from 2026-06-09.
Shared entities / Same source domain / Shared topic / Earlier coverage
Anthropic and Amazon: $25B, 5GW, 1M+ Trainium Chips.
Both cover Amazon, Anthropic; reported by the same outlet (anthropic.com); overlapping topics (amazon, anthropic).
Shared entities / Earlier coverage / Tension
MCP Servers Are The New Shadow IT: 30 CVEs in 60 Days, 82% Vulnerable to Path Traversal, 38% Have Zero Auth
Both cover Anthropic, CVSS, Microsoft; earlier Anthropic coverage from 2026-03-23; pushes against this story (against).
NYT Magazine: "Coding After Coders" Goes Mainstream
Both cover Amazon, Google, Microsoft; earlier Amazon coverage from 2026-03-14; pushes against this story (but).
Source trail
Entities
Claim evidence
- Anthropic, Google, Microsoft, Amazon draft jailbreak severity score
Provenance
- Canonical issue
- 2026-07-02
- AI generated
- yes
- Story unit
- 2026-07-02-anthropic-google-microsoft-and-amazon-are-drafting-a-shared-jailbreak-severity-score
- Labels
- source-backed, canonical briefing excerpt