Data Confidentiality

When using aiExpert to check your models, it is natural to ask: what happens to the data I send to the AI? This article explains how data is handled, and why you can be confident that your information stays secure and private.

How your data is processed

• aiExpert sends selected modelling information (such as element names, relationships, and diagram structures) to a Large Language Model (LLM) – such as GPT-5 (OpenAI) or Gemini (Google).

• The LLM processes the request in real time and returns the feedback immediately.

• No data is stored in the LLM model itself. GPT-5 does not “learn” from your queries and does not use them to train future models.

Storage and retention

• For requests made through your own OpenAI or Gemini account, your queries are transmitted securely over HTTPS and handled according to the provider’s published policies:

  • OpenAI Data Usage Policy

  • Google Gemini Data Privacy

• For requests made through the eaTeamWorks shared trial account, the same safeguards apply. The data is not written to persistent logs and is not accessible to other users.

• Neither aiExpert nor GPT-5 stores your models permanently. Once a response has been generated, the request and response are discarded by the model service.

Logs and monitoring

• System logs may record that a request took place (for billing or quota purposes), but they do not contain the content of your models.

• No other user can see, access, or retrieve the queries you send.

Your control

• You always choose whether to use the built-in eaTeamWorks account (with free trial requests) or your own AI account.

• You can delete or disable stored API keys at any time.

• Feedback returned from the LLM is stored only where you decide: in EA, in eaSheets/Excel, or in documents you generate with eaDocX.