Does my data get used to train an AI model?
No. iDBQuery does not use your data, schemas, or questions to train any AI model. Your data is used only to answer your own questions within your workspace and is never pooled, shared, or fed back into model training. In private-VPC, air-gapped, and Desktop deployments, your data never leaves your environment at all.
Protecting your data from becoming training material is a core commitment, not a setting.
- No training on your data. Your tables, documents, schemas, and the questions you ask are used solely to produce answers for you. They are never used to train or fine-tune any model, and they are never shared across customers.
- Isolation by deployment. In private-VPC, air-gapped, and Desktop deployments, your data physically never leaves your environment, so the question of external training does not even arise.
- Purpose-limited processing. When the AI writes SQL or interprets a document, it works against your data only to fulfill that request, then the result is returned with citations back to the exact source rows.
This matters because many teams are rightly cautious about sending proprietary financials, records, or operational data to AI tools. iDBQuery is designed so that using AI to query your data does not mean surrendering control of it. The combination of in-place querying and data-stays-local deployments means you can adopt natural-language analysis without your information leaking into a shared model.
If you need contractual or architectural assurances for a security review, we are happy to walk your team through exactly how data flows. Reach out and we will provide the detail.
Updated 2026-06-22