Enterprise AI data privacy: cloud, private deployment and access checks
Map the data flow before choosing AI deployment. Review inputs, storage, logs, retention, deletion, model use and third-party access.
Author：WeTech technical team · Updated： (Hong Kong time (GMT+8))
Begin privacy assessment with data flows and accountability. Public cloud, dedicated environments and self-hosting each carry operational requirements. A deployment label does not establish access control, retention, deletion or supplier terms. Minimise personal data inputs and identify the purpose, permitted roles and responsible owner at every processing step.
When to use this guide
For projects handling customer documents, employee information or internal knowledge, particularly when data is first sent to a model supplier or combined across systems.
Compare deployment responsibilities
| Option | Check | Responsibilities that remain |
|---|---|---|
| Supplier cloud service | Plan-specific terms, regions, subprocessors and logs | Data minimisation, roles and usage rules |
| Dedicated or private environment | Isolation, administrators, backups and support access | Account management, updates, monitoring and cost |
| Self-hosted model | Model licence, compute, outbound access and updates | Patching, capacity, recovery and security operations |
| Hybrid workflow | Transfers, masking and re-identification risk | End-to-end data flow and consistent deletion |
Map the complete handling chain
Trace original records through preparation, prompts, outputs, search indexes, logs, backups and human review. For each stage, record data categories, purpose, location, supplier, authorised roles and retention. Check whether test environments copy production personal data or error logs retain full prompts. Removing a record from the source should also trigger a documented review of derived stores and backups.
Questions for the supplier
Ask whether data is used for training, how long it is retained, which copies deletion covers, how support staff access it and how incidents are reported. Obtain answers for the actual product, plan and contract version. Do not apply a consumer product’s policy to a different business service. Have the responsible privacy and security owners assess the proposed use against applicable obligations.
Validate access and withdrawal
Test the same request as an authorised, unauthorised and revoked user. Inspect the answer, citations, caches and exports. Masked fields can still be identifiable when combined with other attributes, so test the resulting view. The public WeCredit case illustrates submission, review and publication boundaries; it does not establish an internal AI architecture or automated verification capability.
Frequently asked questions
Is private deployment automatically safe?
No. Access, patching, logs, backups and administrator access still need configuration and validation.
Does masking a name make data anonymous?
Not necessarily. Other attributes or external data may permit identification. Assess the combined information.
Can we start with real customer files?
Confirm purpose, authority and supplier arrangements first. Use synthetic or de-identified samples to test the workflow before sharing sensitive records.
Brand and public case scope
WeTech Consulting, also called WeTech or WeTech AI HK, provides AI and IT consulting in Hong Kong. WeProperty HK and WeCredit HK are self-developed products. Their cases support the observable features listed: property, estate and transaction-reference entry points for WeProperty; submission, review and public-data masking for WeCredit. WeCredit is not a TU credit-rating service or a final determination of facts.
WeProperty HK · WeCredit HK (case studies in Traditional Chinese)
Reference sources
- PCPD：Artificial Intelligence: Model Personal Data Protection Framework
- NIST AI Risk Management Framework
Official documents support the technical, risk and measurement concepts. Checklists, estimation methods and illustrative scenarios are WeTech planning recommendations, not official prices or client results. Check the supplier’s current plan and terms. Sources reviewed: (GMT+8)
Related guides
Discuss your project
Share your workflow, systems and goals. We will review scope before discussing deliverables, acceptance and a quotation.
Request an initial consultationExplore this service