Google Vertex AI data use review for SaaS customer commitments
Vertex AI data-use answers should cite Google Cloud's Vertex AI data governance source and, where relevant, the separate zero data retention source. The customer answer should name whether the workflow uses Vertex AI, Gemini through Google Cloud, tuning, feedback, or customer-controlled logging.
Vendor
Google Vertex AI / Gemini for Cloud
Issue
data use
Sources reviewed
3 official sources
Product and plan applicability
- Scope
- Vertex AI generative AI
- Applies to
- Google Cloud-hosted prompts, responses, files, embeddings, and model calls.
- Watch for
- Model, endpoint, region, tuning, feedback, safety logging, and project-level storage.
- Scope
- Zero data retention
- Applies to
- Supported Vertex AI generative AI requests described by Google.
- Watch for
- Feature and model eligibility before promising zero retention to customers.
- Scope
- Google Workspace Gemini
- Applies to
- Employee productivity tools, not Vertex AI product features.
- Watch for
- Use Workspace sources rather than Vertex AI sources.
| Scope | Applies to | Watch for |
|---|---|---|
| Vertex AI generative AI | Google Cloud-hosted prompts, responses, files, embeddings, and model calls. | Model, endpoint, region, tuning, feedback, safety logging, and project-level storage. |
| Zero data retention | Supported Vertex AI generative AI requests described by Google. | Feature and model eligibility before promising zero retention to customers. |
| Google Workspace Gemini | Employee productivity tools, not Vertex AI product features. | Use Workspace sources rather than Vertex AI sources. |
What official sources say
Data governance is the main source
Google Cloud's Vertex AI data governance page is the primary source for how prompts, responses, and customer data are handled for Vertex AI generative AI.
Zero retention needs its own eligibility check
Google publishes a separate Vertex AI zero data retention page. Treat zero retention as a feature-specific statement, not a blanket Google Cloud claim.
Why a SaaS team should review it
- Vertex AI commitments may differ from Workspace Gemini, consumer Gemini, or other Google services.
- Teams may store prompts and responses in their own Google Cloud logs even when model-side handling is limited.
- Customer security reviews often ask for source links that show training, retention, and subprocessor treatment.
Potential customer commitment drift
- A Trust Center says Google does not retain AI prompts, but the product stores prompts in BigQuery or Cloud Logging.
- A feature moves from standard Vertex AI calls to tuning or feedback flows without a source review.
- Sales uses zero retention language before confirming supported model and endpoint coverage.
Review checklist
- Identify the exact Google Cloud AI service, model, endpoint, and project.
- Check data governance and zero retention sources before customer publication.
- Inventory Cloud Logging, BigQuery, app databases, and support tickets for prompt or output copies.
- Review Google Cloud DPA and terms if personal data is involved.
- Record source links and last reviewed date with the customer answer.
Source links
Sources were reviewed on 2026-05-21. This page supports a review packet or monitoring evidence packet; it is not legal advice.
Related pages
Scan Google Vertex AI / Gemini for Cloud against your own commitments.
Use this page as a starting point, then compare the vendor source to the exact promise in your Trust Center, DPA, security questionnaire, or sales answer. The $199 packet turns that review into cited evidence your team can route internally.