canonical: https://jentic.com/apis/googleapis.com/bigquerydatapolicy

# Google BigQuery Data Policy API

The BigQuery Data Policy API manages column-level masking policies that BigQuery enforces at query time. You attach a data policy to a policy tag in Data Catalog, and any column tagged with it is masked, hashed, or nulled out for callers who lack the unmask role. The API exposes create, list, patch, rename, and delete operations on data policies plus IAM policy management on each resource. Together with policy tags, this lets data platform teams enforce GDPR/HIPAA-style column controls without writing per-query views.

## For AI agents

Manage column-level masking and access policies attached to BigQuery policy tags. Agents can create, rename, and IAM-bind data policies that protect sensitive columns at query time.

## Scope

Does not run queries, manage encryption keys, or transfer data - use for column-level masking policy management only.

## Capabilities

- Create data policies that apply masking rules to columns tagged with a policy tag
- List all data policies in a project and location for inventory and audit
- Patch a data policy to change its data-masking expression
- Rename a data policy resource without re-tagging columns
- Delete data policies that are no longer enforced
- Read and set IAM policies that grant the unmask role on sensitive columns
- Test caller permissions on a data policy before delegating access

## Use cases

### PII Masking at Query Time

Apply masking policies to columns tagged with sensitivity labels so PII is hashed or nulled for callers without the unmask role. Data policies are managed at /v1/{parent}/dataPolicies and bound via setIamPolicy. This avoids creating per-team views and keeps a single source of truth for the underlying table.

Example prompt: Create a data policy named mask_email_sha256 that applies SHA256 to columns tagged with the pii.email policy tag in project myproj location US.

### Tiered Access Across Analyst Groups

Grant the dataPolicies.maskedReader role to broad analyst groups and roles/datacatalog.categoryFineGrainedReader to a small unmask group. Use setIamPolicy on the data policy resource to control which principals see the raw values. The whole tiering can be expressed as a few API calls, avoiding hand-managed BigQuery views.

Example prompt: Bind roles/bigquerydatapolicy.maskedReader to group analysts@example.com on data policy projects/myproj/locations/US/dataPolicies/mask_email_sha256.

### Compliance Audit and Inventory

Periodically list all data policies in each project and location and verify that critical policy tags have a policy attached. The list endpoint at /v1/{parent}/dataPolicies returns the policy expression and bound resources, which can be exported into a compliance system.

Example prompt: List all data policies in project myproj location US and return any that have an empty data masking expression.

### Agent-Driven Policy Provisioning

An AI agent can take a request like 'mask credit card numbers everywhere they appear' and create the matching data policy via Jentic. The agent loads the create operation, generates the masking expression, and binds the right IAM role to a reviewer group. Sensitive credentials never leave your Jentic One instance.

Example prompt: From the request 'redact CCN columns for everyone except finance', create the data policy and grant unmask role to finance-readers@example.com.

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | /v1/{+parent}/dataPolicies | Create a new data policy under a project and location |
| GET | /v1/{+parent}/dataPolicies | List data policies under a project and location |
| GET | /v1/{+name} | Get a single data policy |
| PATCH | /v1/{+name} | Update a data policy's masking expression |
| POST | /v1/{+name}:rename | Rename a data policy resource |
| DELETE | /v1/{+name} | Delete a data policy |
| POST | /v1/{+resource}:setIamPolicy | Set the IAM policy on a data policy |

## Key resources

- **Data Policies** — Create, rename, patch, list, and delete data masking policies attached to policy tags.
- **IAM Policies** — Manage who can view masked vs unmasked values for tagged columns.

## Why Jentic

- **Setup:** Wiring the BigQuery Data Policy API by hand means setting up Google OAuth2, scoping a service account for policy management, and minting short-lived access tokens for each masking policy call yourself. Through Jentic you install once, import the BigQuery Data Policy API from the API Directory, store the OAuth credential once, and your agent calls it.
- **Permission scoping:** This API puts the data policy resource name in the URL path (/v1/{parent}/dataPolicies, /v1/{name}), so a rule can pin your agent to one project or policy: it can create, list, and read data policies there and nothing else. You choose the operations it may call, so actions like deleting a policy or setting its IAM policy are not included unless you add them.
- **Credential handling:** Your BigQuery OAuth credential is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'mask a BigQuery column', and Jentic returns the create dataPolicy operation with its dataMaskingPolicy schema so the agent calls the right endpoint without reading the discovery document.

## Related APIs

- **BigQuery API** — Runs the queries that the data policies mask at execution time.
- **Cloud KMS API** — Manages encryption keys that protect the underlying data at rest.
- **BigQuery Connection API** — Provides a different control plane for federated source access.

## FAQ

### What authentication does the BigQuery Data Policy API use?

The BigQuery Data Policy API uses Google OAuth 2.0 with the cloud-platform and bigquery scopes. Jentic stores the OAuth refresh token encrypted and provides the agent only a scoped access token at execution time.

### Can I attach a masking expression to a Data Catalog policy tag?

Yes. Create the data policy via POST /v1/{parent}/dataPolicies with a policyTag field referencing the Data Catalog tag and a dataMaskingPolicy expression (for example SHA256, DEFAULT_MASKING_VALUE, or a custom UDF). BigQuery enforces it on every query.

### What are the rate limits for the BigQuery Data Policy API?

Project-level quotas apply: typically a small write QPS for create/patch/rename/delete on the same data policy and a few hundred QPS for reads. Specific per-project values are visible in the Google Cloud quotas console.

### How do I rename a data policy through Jentic?

Search Jentic for 'rename a BigQuery data policy', load the schema for POST /v1/{name}:rename, and execute with the new newDataPolicyId. Jentic forwards the OAuth token; the policyTag binding stays intact.

### Does this API mask data in tables that have no policy tag?

No. Data policies only take effect on columns whose schema includes a policyTag that this policy is attached to. Untagged columns are returned in cleartext regardless of how many data policies exist.

### Can I limit what my agent is allowed to do with the BigQuery Data Policy API?

Yes. Because your Jentic One instance is self-hosted, your own rules decide which operations and credentials the agent may use. Since this API puts the data policy resource name in the URL path (/v1/{parent}/dataPolicies and /v1/{name}), you can pin the agent to a single project or policy and let it only create, list, and read data policies there. Destructive or access-granting calls like deleting a policy or running setIamPolicy stay off limits unless you explicitly add them.