Skip to content
Closed
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
21 changes: 15 additions & 6 deletions content/en/security/cloud_security_management/_index.md
Original file line number Diff line number Diff line change
Expand Up @@ -81,16 +81,24 @@ Cloud Security also includes access to Datadog Security features, including:
- [Audit Trail][20]
- [Security Research Feed][16]

{{< callout url="#" btn_hidden="true" header="false">}}
The new Cloud Security Summary shown below is in Preview. To get started, contact your Customer Success representative.
{{< /callout >}}

{{< img src="security/csm/csm_overview_4.png" alt="Cloud Security Summary in Datadog" width="100%">}}
{{< img src="security/csm/csm_overview_5.png" alt="Cloud Security Summary in Datadog" width="100%">}}

{{< partial name="security-platform/CSW-billing-note.html" >}}

## Track your organization's health

### Manage homepage dashboards

You can customize the dashboards that you can access directly from the Cloud Security homepage, including setting a dashboard to your default homepage view. Use dashboards to prioritize remediation efforts, schedule reports, put security data next to observability and cost data, and embed apps and workflows you can start directly from your reporting view.

On the [Cloud Security homepage][4], in the **Dashboards** section, you can jump directly to dashboards for Identity Risks, Misconfigurations, or Vulnerabilities. You can add existing dashboards or create one to keep on your Cloud Security homepage sidebar for convenient access.

Additionally, you can click the **More Options** icon {{< img src="icons/kebab.png" inline="true" style="height:1em" >}} to manage your pinned dashboards, including setting one as your Cloud Security homepage default view. Click **Cloud Security** in the Datadog nav bar or **Summary** in the Cloud Security nav bar to go directly to your pinned dashboard.

For more information, see [Dashboards][23].

### Track your security posture score

Available for [Cloud Security Misconfigurations][2], the [security posture score][5] helps you track your organization's overall health. The score represents the percentage of your environment that satisfies all of your active out-of-the-box cloud and infrastructure compliance rules.

Improve your organization's score by remediating misconfigurations, either by resolving the underlying issue or by muting the misconfiguration.
Expand Down Expand Up @@ -153,4 +161,5 @@ To get started with Cloud Security, navigate to the [**Cloud Security Setup**][3
[19]: /security/automation_pipelines/
[20]: /security/audit_trail/
[21]: https://app.datadoghq.com/security/map
[22]: /datadog_cloudcraft/overlays/#security
[22]: /datadog_cloudcraft/overlays/#security
[23]: /dashboards/
Original file line number Diff line number Diff line change
@@ -0,0 +1,128 @@
---
title: Ownership Agent
further_reading:
- link: "/security/cloud_security_management/guide/frontier_group/ownership_preferences"
tag: "Documentation"
text: "Set Up Ownership Preferences"
- link: "/security/cloud_security_management/guide/frontier_group"
tag: "Documentation"
text: "Cloud Security Frontier Group"
- link: "/security/cloud_security_management/misconfigurations"
tag: "Documentation"
text: "Cloud Security Misconfigurations"
---

## Overview

The Ownership Agent analyzes cloud resources that have security findings and infers who is responsible for each one. For every resource, it produces a suggested owner—a team or individual—along with a confidence score, a short explanation, and a list of the evidence signals it used.

Ownership suggestions appear in the Cloud Security side panel when you view a misconfiguration or finding. You can confirm, reject, or correct each suggestion, and your feedback improves future inference results.

## Data sources

The Ownership Agent reads multiple data sources and combines them into a ranked evidence set. It evaluates data sources in priority order: stronger, more explicit signals override weaker or inferred ones.

| Priority | Signal | Description |
| --- | --- | --- |
| 1 | **Owner tags** | Cloud resource tags with keys such as `owner`, `dd-team`, or `team`. An explicit ownership tag is the strongest signal. |
| 2 | **Ownership preferences** | Custom tag mappings and rules you define in a [reference table][1]. These act as organization-level overrides and are evaluated alongside direct tags. |
| 3 | **Service Catalog** | Team ownership data from the Datadog Service Catalog, matched against the resource's service, application, or component tags. |
| 4 | **Cloud audit logs** | The identity of the user or principal that created the resource, extracted from cloud provider audit logs (for example, AWS CloudTrail). Automation accounts and CI principals are filtered out. |
| 5 | **IaC last committer** | The most recent human author to commit infrastructure-as-code that provisioned or modified the resource, when available from source metadata. |
| 6 | **Container and host catalog** | Registry and host metadata for container images and host VMs, including image labels and host annotations. |
| 7 | **Naming patterns** | Heuristics that infer ownership from resource names, service identifiers, or tag values that match known team naming conventions. |

When deterministic signals are strong enough, the Ownership Agent resolves ownership without using AI. For ambiguous cases—where signals conflict or are missing—the agent uses an AI model to synthesize a suggestion from the available evidence. The AI model is governed by a per-organization token budget and falls back to heuristic output if the budget is exhausted.

## Supported resource types

The Ownership Agent processes the following cloud resource types:

| Resource type | Cloud provider |
| --- | --- |
| `aws_ec2_instance` | AWS |
| `aws_s3_bucket` | AWS |
| `host` | AWS, Azure, GCP |
| `host_image` | AWS, Azure, GCP |
| `image` (container) | Docker, ECR, and other registries |

The agent runs on a 2-hour cadence for resource evidence and a 24-hour cadence for team data. Coverage is limited to resources with at least one active security finding.

## Review and correct ownership

### View a suggestion

When you open a finding in the [Misconfigurations Explorer][2], the side panel displays the suggested owner under **Ownership**. Each suggestion includes:

- The suggested owner handle and type (team, user, or service)
- A confidence score from 0 to 1
- A one- to two-sentence explanation of why the agent chose that owner
- The evidence signals that contributed to the result

### Confirm, reject, or correct

You can respond to any suggestion directly in the side panel:

| Action | What it does |
| --- | --- |
| **Confirm** | Marks the suggestion as accurate. The suggestion remains in _suggested_ status and the positive signal is recorded for future tuning. |
| **Reject** | Marks the suggestion as inaccurate. The suggestion remains in _suggested_ status and the negative signal is recorded for future tuning. |
| **Correct** | Replaces the suggested owner with the owner you specify. The status transitions to _overridden_ and the correction is stored in the audit history. |

### Persist ownership

A suggestion is non-authoritative by default. To lock in an owner for downstream workflows and assignment:

- Click **Persist** to explicitly save the current suggestion. The status transitions to _persisted_.
- Your organization can opt in to automatic persistence when confidence is at or above 0.85. Contact [Datadog Support][3] to enable this option.

Persisted ownership is used for routing, assignment, and reporting. Overridden ownership (from a manual correction) is treated the same as persisted.

### Impact on evaluation

Corrections and persistence affect how the Ownership Agent behaves for that resource going forward:

- **Overridden** resources retain your correction across re-evaluation cycles. If evidence changes significantly, the agent flags the discrepancy but does not automatically overwrite your correction.
- **Persisted** resources are re-evaluated when the underlying evidence changes (for example, if a resource's tags change). You are notified if the new inference differs from the persisted value.
- **Confirmed and rejected** signals accumulate as training feedback. They are used to tune heuristic weights and improve the accuracy of future suggestions, but do not change the current suggestion's status.

## Query ownership in the explorer

You can filter findings by inferred owner in the [Misconfigurations Explorer][2] using the following facets:

| Facet | Description | Example |
| --- | --- | --- |
| `Ownership > Owner` | The suggested or persisted owner handle | `@ownership.owner_handle:team-platform` |
| `Ownership > Owner type` | The type of owner: `team`, `user`, or `service` | `@ownership.owner_type:team` |
| `Ownership > Status` | The lifecycle status of the suggestion | `@ownership.status:persisted` |
| `Ownership > Confidence` | The numeric confidence score (0–1) | `@ownership.confidence:>0.8` |

**Example queries**

To find all findings owned by a specific team:

```
@ownership.owner_handle:team-platform
```

To find high-confidence suggestions that have not yet been reviewed:

```
@ownership.status:suggested @ownership.confidence:>0.85
```

To find findings with a manually corrected owner:

```
@ownership.status:overridden
```

Ownership facets can be combined with any other explorer filter, such as resource type, cloud account, or rule severity.

## Further reading

{{< partial name="whats-next/whats-next.html" >}}

[1]: /security/cloud_security_management/guide/frontier_group/ownership_preferences/
[2]: https://app.datadoghq.com/security/compliance
[3]: /help/
Original file line number Diff line number Diff line change
Expand Up @@ -49,6 +49,14 @@ Select a rule to view details about the misconfigured resources, the rule descri

{{< img src="security/cspm/frameworks_and_benchmarks/failed-finding3.png" alt="The compliance rule side panel includes information about the rule and resources with failed misconfigurations" style="width:75%;">}}

## Track compliance posture with dashboards

You can use dashboards to see posture scores across teams, services, accounts, environments, and more. This data can help you prioritize your remediation efforts, schedule reports, put security data next to observability and cost data, and embed apps and workflows so you can start them directly from your reporting view. For more information, see [Dashboards][4].

You can graph your organization's compliance data in dashboards using the `datadog.csm.findings.count` metric. You can also see your organization's [security posture score][3] in the OOTB Compliance Dashboard.

<div class="alert alert-info">This metric estimates the true values in the Compliance page, and some error is expected. You can filter the <a href="https://app.datadoghq.com/security/compliance/home/essential-cloud-security-controls">Compliance Framework page</a> to see exact values for any combination of account, service, environment, and team.</div>

## Show or hide compliance frameworks

You can hide frameworks from the list on the [Compliance][20] page, so you can focus on the ones that are most relevant to your organization.
Expand All @@ -64,6 +72,7 @@ Then, on the Compliance page, you can use the **Show hidden frameworks** toggle
[1]: /security_monitoring/default_rules/
[2]: /security/cloud_security_management/misconfigurations/frameworks_and_benchmarks/supported_frameworks/
[3]: /glossary/#security-posture-score
[4]: /dashboards
[11]: /security/cloud_siem/
[12]: /integrations/amazon_cloudtrail/
[13]: https://app.datadoghq.com/security/configuration/rules?product=cspm
Expand Down
2 changes: 1 addition & 1 deletion layouts/partials/security-platform/CSW-billing-note.html
Original file line number Diff line number Diff line change
@@ -1 +1 @@
<div class="alert alert-info">Collecting events using Cloud Security Management will affect your billing. For more information, see <a href="https://www.datadoghq.com/pricing/?product=cloud-security-management#products">Datadog Pricing</a>.</div>
<div class="alert alert-info">Collecting events using Cloud Security affects your billing. For more information, see <a href="https://www.datadoghq.com/pricing/?product=cloud-security-management#products">Datadog Pricing</a>.</div>
Binary file added static/images/security/csm/csm_overview_5.png
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading