# Agent 365 data connector is enabled on the Microsoft Sentinel workspace

The Agent 365 data connector brings AI agent telemetry into Microsoft Sentinel, including agent invocations, tool calls, and grounding source access.

Source: https://maester.dev/docs/tests/ZTA.61023

<!-- This file is generated by website/scripts/generate-test-docs.mjs. Do not edit manually. -->

## Overview

The Agent 365 data connector brings AI agent telemetry into Microsoft Sentinel, including agent invocations, tool calls, and grounding source access. Without it, there is no way to see whether an AI agent was involved in a session, what it did, or what data it accessed. This check verifies the connector is installed and connected on at least one Sentinel-onboarded workspace.

When Agent 365 telemetry is not forwarded to Sentinel, the workspace has no visibility into what AI agents ran, what tools they invoked, or what data they accessed. Threat actors who manipulate an agent through prompt injection can exploit this blind spot because their entire footprint — tool invocations, grounding source queries, and induced outputs — exists only in the agent execution layer, which is invisible to the SIEM. Without this connector, the organization cannot distinguish whether a data access event was performed by a human user or by an agent operating under attacker control.

## Remediation

- [A365 Observability solution for Sentinel — Marketplace listing](https://marketplace.microsoft.com/en-us/product/saas/azuresentinel.azure-sentinel-solution-a365observability)
- [A365 Observability GitHub source and release notes](https://github.com/Azure/Azure-Sentinel/tree/master/Solutions/A365%20Observability)
- [Install solutions from the Microsoft Sentinel Content Hub](https://learn.microsoft.com/azure/sentinel/sentinel-solutions-deploy)
- [Microsoft Sentinel data connectors overview](https://learn.microsoft.com/azure/sentinel/connect-data-sources)

## Run it

The check runs as part of the whole assessment, with the ZeroTrustAssessment module installed:

```powershell
Install-Module ZeroTrustAssessment -Scope CurrentUser
Connect-Maester
Invoke-Maester -Pack ZTA -SkipBuiltIn
```

Leave out `-SkipBuiltIn` to run Maester's own checks in the same run.

<details>
<summary>About this check</summary>

| | |
| --- | --- |
| Severity | High |
| Effort to fix | Low |
| Impact on users | Low |
| Pillar | AI |
| Category | AI Threat Detection |
| Product | [Azure](/docs/tests/products/azure) |
| Feature | Microsoft Sentinel |
| Area | Logging and detection |
| Licence | AGENT_365, Consumption-based: Microsoft Sentinel |
| Tenant | Workforce |

Product, feature, area and ATT&CK are Maester's: the assessment does not record them, so Maester keeps them in the pack's `taxonomy.json`.

</details>

<p class="mt-ms-note not-content"><svg class="mt-ms-note__logo" viewBox="0 0 21 21" aria-hidden="true"><rect width="10" height="10" fill="#f25022"/><rect x="11" width="10" height="10" fill="#7fba00"/><rect y="11" width="10" height="10" fill="#00a4ef"/><rect x="11" y="11" width="10" height="10" fill="#ffb900"/></svg><span>A Microsoft check, part of the <a href="/docs/tests/packs/zero-trust-assessment">Zero Trust Assessment</a> pack. Text from ZeroTrustAssessment 3.1.0, © Microsoft, <a href="/docs/tests/packs/zero-trust-assessment#licence">MIT licence</a>.</span></p>
