All your data, including the audio files of your conversations, is not used to train the AI.
This is excluded from training datasets. In other words, OpenAI does not train its models on your conversation data or the insights (summaries, topics, sentiments) derived from it.
Furthermore, your data is not stored longer than necessary for processing.
For the purpose of improving our product and gaining insight into the use of the platform, Rinkel does use generic data, but only in large numbers (bulk) and not specific to individual customers. In other words, we do not monitor how AI insights is used by each customer or company, but only to gain global insights into usage in general.