# Training on our Conversations?

**URL:** https://community.heypocket.com/t/training-on-our-conversations/862
**Category:** General
**Created:** [November 13, 2025, 6:18pm UTC](https://community.heypocket.com/t/training-on-our-conversations/862 "2025-11-13T18:18:43Z")
**Posts on this page:** 2
**Page:** 1

<div class="post-metadata">

### Author: ![dps](https://yyz1.discourse-cdn.com/flex029/user_avatar/community.heypocket.com/dps/32/107_2.png) [@dps](https://community.heypocket.com/u/dps)
#### Post date: [November 13, 2025, 6:18pm UTC](https://community.heypocket.com/t/training-on-our-conversations/862/1 "2025-11-13T18:18:43Z")

</div>

Haven’t seen this discussed much, but I am wondering to what degree our data with Pocket can and/or is used to train models. Both external provider models and any internal models. I would appreciate having the option to disable any and all training, as well as confirmation on how long providers may have our information for and in what capacity.

---

<div class="post-metadata">

### Author: ![DaveNgl](https://avatars.discourse-cdn.com/v4/letter/d/b77776/32.png) [@DaveNgl](https://community.heypocket.com/u/DaveNgl)
#### Post date: [November 14, 2025, 6:04pm UTC](https://community.heypocket.com/t/training-on-our-conversations/862/2 "2025-11-14T18:04:41Z")

</div>

They advertised that it would not be used for any training, and I don’t see how they could allow it and still be HIPPA compliant.
