---
id: "CVE-2026-119344"
kind: "vulnerability"
aliases: ["CVE-2026-119344"]
title: "Cocos AI: attestation accepts evidence from other platforms, letting an attacker be authorized"
severity: "critical"
cvss: 9.1
epss: null
exploited: false
status: "open"
category: "other"
products: []
ecosystems: []
tags: ["attestation", "cocos-ai", "confidential-computing", "other", "tee"]
not_affected_if: ["not-using-cocos-ai", "version-0.9.0-or-later"]
published_at: "2026-11-05T16:17:40+00:00"
first_seen_at: "2026-11-05T16:40:01+00:00"
updated_at: "2026-11-05T16:41:20+00:00"
sources: [{"source": "nvd", "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-119344", "first_seen_at": "2026-11-05T16:40:01+00:00"}]
exploit_sources: []
---

# Cocos AI: attestation accepts evidence from other platforms, letting an attacker be authorized

Cocos AI, a confidential computing system, accepts trusted execution environment attestation evidence without binding it to a specific platform owner or machine identity, and does not enforce revocation. An attacker who controls another platform trusted by the same verifier can substitute that platform's valid evidence and get the wrong platform authorized.

_Summary, check and fix written by a model from the sources; labelled as such._

## How to check

- Check the installed Cocos AI version; versions `0.8.2` and earlier are affected.

## Fix

- Move to `0.9.0`, which the description says addresses this issue.
- The description notes this cannot be fully fixed in source code alone because it stems from the threat model; additional deployment-level measures are not stated in the sources.

## Timeline

- 2026-11-05T16:40:01+00:00 nvd https://nvd.nist.gov/vuln/detail/CVE-2026-119344

## Documents

- [advisory] Vendor advisory: github.com https://github.com/ultravioletrs/cocos/security/advisories/GHSA-v5m8-5wxc-vjgp
- [fix] Release notes: github.com https://github.com/ultravioletrs/cocos/releases/tag/v0.9.0

## From the sources

Cocos AI is a confidential computing system for AI. In version 0.8.2 and earlier, Cocos AI accepts trusted execution environment attestation Evidence when firmware patch levels and software measurements match, without binding the Evidence to an allowed platform owner or unique machine identity and without enforcing revocation. A remote attacker who controls or compromises another platform trusted by the same verifier can substitute valid Evidence from that platform and cause authorization of the wrong platform. This vulnerability is addressed in 0.9.0. However, this vulnerability can't be patched in source code alone, since the vulnerability arises from the confidential computing threat model itself.

---

_Free to read, with fair use of about 60 pages a minute. For higher limits or API access: sales@infrastack.ai_