> For clean Markdown of this page, append .md to its URL. For the complete documentation index, see https://www.superagent.sh/llms.txt.


Check whether an AI agent respects the boundary between customer workspaces.

# Tenant isolation for AI agents


[← Back to attack scenarios](https://www.superagent.sh/docs/products/red-teaming/attack-scenarios)

A shared agent service can serve many customers while keeping their data and actions separate. This scenario checks whether the agent stays inside the active customer's workspace when it reads information or acts on someone's behalf.

It applies to agents with a tenant or organization boundary. A finding identifies the workspace boundary and the kind of access or action it failed to contain.

The app lists this scenario as **Cross a customer or tenant boundary** (`AG-02`) in [Agent Red Teaming](https://www.superagent.sh/docs/products/red-teaming/agents).

## Standards and guidance

[OWASP Agentic Top 10 (2026), ASI03 Identity & Privilege Abuse](https://genai.owasp.org/resource/owasp-top-10-for-agentic-applications-for-2026/)

---
Source: https://www.superagent.sh/docs/products/red-teaming/attack-scenarios/agent-tenant-isolation
Index: https://www.superagent.sh/llms.txt
