Excessive agency / agent goal hijack
Over-privileged agents redirected to attacker goals
categoryAI/ML and emerging attacksseverityHighlabnot yet measured
Classification
18 values: 7 imported, 1 derived, 10 machine-classified. What do these mean?
Exploitability
| Attack vector | AI model manipulation |
|---|---|
| Reach (CVSS) | Network |
| Attack complexity | Low |
| Privileges required | None |
| User interaction | Passive |
| Attacker skill required | Low |
| Resources required | Low |
| Likelihood / prevalence | High |
| Maturity | Emerging |
Impact & scope
| Severity | High |
|---|---|
| Impact | ConfidentialityIntegrity |
| Platform | AI System |
| Environment | AI system |
| Target asset class | ML model / agent |
| Kill-chain phase (tactic) | ExecutionLateral MovementPrivilege Escalation |
Detection & response
Mitigated by 13
- AML.M0019 atlas-mitigation
- AML.M0020 atlas-mitigation
- AML.M0021 atlas-mitigation
- AML.M0022 atlas-mitigation
- AML.M0024 atlas-mitigation
- AML.M0026 atlas-mitigation
- AML.M0027 atlas-mitigation
- AML.M0028 atlas-mitigation
- AML.M0029 atlas-mitigation
- AML.M0030 atlas-mitigation
- AML.M0032 atlas-mitigation
- AML.M0033 atlas-mitigation
- AML.M0035 atlas-mitigation
Framework references
- Primary source: https://goteleport.com/blog/owasp-top-10-agentic-applications/
- owasp-llm: LLM06:2025 (primary)
- owasp-asi: ASI01
- atlas: AML.T0053
- atlas: AML.T0051
Evidence & sources
- ev-0001 (Cyber Attack Taxonomy) — Name, definition and Example/ID taken verbatim from the 'AI/ML and emerging attacks' table, line 628 of Cyber-Attack-Taxonomy.md
- ev-0002 (Seeded-record enrichment run 2026-09-30 (crosswalk + facet classification)) — Crosswalk to atlas:AML.T0053, atlas:AML.T0051 as the nearest framework entries; ids verified to exist in the cached catalogues.