If you do not know what the platform keeps, you do not know the real risk
Ask a vendor what it stores from prompts, uploads, outputs and logs, and for how long. Check the settings and contract for the exact product tier, including who can access retained data and how deletion works.
If prompts, uploaded files, outputs or metadata are stored longer than expected, review the implications for confidentiality, compliance and client contracts.
Include sensitive commercial, personal and client information in that review, using examples from the firm's actual work.
Assume some retention exists until the vendor proves otherwise
Ask whether the product stores inputs for logging, security, performance monitoring, abuse detection, troubleshooting or product improvement. If the vendor offers reduced-retention or no-training settings, check what each setting covers and whether it is enabled for the intended account.
Confirm what is retained in the environment the firm will use and which controls apply. A conversational interface does not establish how the underlying records are handled.
Buyers should ask:
- Are prompts, uploads, outputs, and metadata retained?
- What are the default retention periods?
- Can retention be shortened, disabled, or configured by policy?
- Who inside the vendor can access retained records?
- Do subprocessors or model providers also retain any data?
- How are deletion requests and legal holds handled?
Keep unanswered questions in the supplier review and resolve them before approving sensitive data for that use.
Check retention risk before AI adoption
What counts as user input retention in practice
Review retention across the records associated with a request, including the prompt and any related files, outputs or logs.
For example, a platform may retain:
- prompt text entered by users
- uploaded files or extracted document content
- model outputs and conversation history
- usage logs, timestamps, account identifiers, and IP data
- feedback signals used for product improvement or safety review
- audit and support records linked to the interaction
When a vendor describes retention as limited, ask which of these categories the statement covers and which have separate storage or deletion rules.
Why retention matters beyond privacy policy language
Use the retention review to decide what data the proposed workflow may process.
If inputs are retained, firms need to think about:
- Confidentiality: whether client or internal information could remain accessible longer than intended.
- Compliance: whether retention conflicts with data minimisation, record handling, or sector obligations.
- Security exposure: whether stored prompt history creates a larger attack surface.
- Contract risk: whether client agreements restrict storage or downstream access.
- Operational discipline: whether teams need stronger prompt hygiene, redaction, or usage limits.
Record the permitted data, chosen settings and staff instructions alongside the supplier's retention answers.
Design governance around AI data handling
The warning signs buyers should notice
Check vendor answers for these gaps:
- claims that the vendor does not retain data, followed by exceptions buried in documentation
- different retention rules across consumer, team, and enterprise products
- unclear answers on metadata, logs, or support access
- no distinction between training use and operational retention
- no practical explanation of deletion and access control
- sales language that is more confident than the contract terms
Resolve these gaps before rollout and record any use restrictions that remain.
How firms should handle this during implementation
Set internal use rules alongside the supplier review, including the data and accounts staff may use.
Before rollout, firms should decide:
- what categories of data are allowed into the platform
- when anonymisation or redaction is mandatory
- which users or teams can access higher-risk features
- what approval or review steps apply to sensitive use cases
- how retention settings are verified and monitored over time
- what fallback process exists when data cannot be entered safely
Assign an owner to verify the chosen settings and keep the use rules current.
A simple buyer rule
Before approving a sensitive workflow, record what the product retains, for how long and who can access it. Use that record to set the data boundary and staff instructions, keeping unresolved questions visible to the approval owner.
Where retention remains unclear, firms should narrow the use case, strengthen review requirements, or pause deployment until the control position is clearer.
Turn retention answers into practical workflow controls
Conclusion
Check storage, duration, access and deletion for the exact product tier, account and settings under review. Compare the answers with the firm's confidentiality and record-handling requirements before approving the proposed use.
Keep that use unapproved while material retention questions remain unresolved.
Need a safer AI route for your firm?
For the proposed tool, record the permitted data, retention settings, deletion process and approval owner. Resolve missing supplier answers before introducing sensitive work.
