risk
Unauthorized or unsafe autonomous agent actions and tool calls
AI agents with excessive permissions or weak action controls execute tool calls, transactions, or code outside their authorized task scope — through prompt injection, misinterpretation, or emergent behaviour — causing data loss, financial loss, or irreversible changes in connected systems.
In catalog since 2026-09-17T22:28:00Z · Last changed 2026-09-17T22:28:00Z (f368a6cce277)
Record JSON · Open in map · Data retrieval guide
Catalog revision: 791ff2dd3a45707290badee660f185e514d15f1cf518908628f425c2f2c56ee4. A connection does not establish full coverage.
Attributes
- category
- ai_governance
- domain
- AI Governance
- Access Control & Identity Management
- taxonomy
- owasp-llm-top10-2025
- iso-23894-ai-risk
- inherent_rating
- high
Details
- risk_id
- ai-agent-unauthorized-actions
- category
- ai_governance
- likelihood
- high
- impact
- high
- inherent_rating
- high
- treatment
- mitigate
- taxonomies
- owasp-llm-top10-2025
- iso-23894-ai-risk
Source
No record-specific source URL is provided.
Connections
- UC-AI-19 — Constrain agent actions and tool use to authorized scope mitigates Unauthorized or unsafe autonomous agent actions and tool calls
- strength
- primary
- rationale
- Tool allow-lists, task-scoped permissions, approval gates for irreversible actions, and sandboxed execution are the direct inverse of excessive agency.
- UC-AI-16 — Ensure human oversight of AI decisions mitigates Unauthorized or unsafe autonomous agent actions and tool calls
- strength
- related
- rationale
- Human overseers who can halt or override the system are the backstop when an agent acts outside its intended scope.
- UC-AI-21 — Commission independent third-party AI evaluations on a quarterly cadence mitigates Unauthorized or unsafe autonomous agent actions and tool calls
- strength
- related
- rationale
- Independent testing of tool calls surfaces unsafe or unauthorized agent actions before customers do.