Zero-click injection via documents/email
Injected text in auto-processed items exfiltrates data with no click
categoryAI/ML and emerging attacksseverityHighlabnot yet measured
Example: EchoLeak
Classification
14 values: 6 reviewed, 6 imported, 2 derived. What do these mean?
Exploitability
| Attack vector | AI model manipulationSoftware vulnerabilities |
|---|---|
| Reach (CVSS) | Network |
| Attack complexity | Low |
| Privileges required | None |
| User interaction | None |
| Attacker skill required | Medium |
| Resources required | Low |
| Likelihood / prevalence | Medium |
| Maturity | Proof of concept |
Impact & scope
| Severity | High |
|---|---|
| Impact | Confidentiality |
| Environment | AI system |
| Target asset class | ML model / agent |
Detection & response
No linked mitigations or detections yet.
Framework references
- Primary source: https://www.hackthebox.com/blog/cve-2025-32711-echoleak-copilot-vulnerability
- cve: CVE-2025-32711 (primary)
Evidence & sources
- ev-0001 (Cyber Attack Taxonomy) — Name, definition and Example/ID taken verbatim from the 'AI/ML and emerging attacks' table, line 607 of Cyber-Attack-Taxonomy.md
- ev-0002 (FIRST EPSS scores) — FIRST EPSS for CVE-2025-32711: epss=0.08034 percentile=0.94510 (score date 2026-09-18).
- ev-0003 (NVD CVE API 2.0) — NVD maps CVE-2025-32711 to weakness(es) CWE-74.