This might be written somewhere, but you have to trust those words.

Most big players in the LLM field are getting their training data by at-least shady, if not illegal measures.

And if they don't care about laws on one side, why should anyone believe that they care on the other?

Amazon already uses their customers privat data to train models [1].

[1] https://www.ftc.gov/business-guidance/blog/2023/06/hey-alexa...

But you have to consider that AWS with all their enterprise customers is something entirely different than Alexa.

And yes, it’s literally written in their documentation.

https://aws.amazon.com/q/developer/faqs/#topic-1

“ For users who access Amazon Q Developer with the Pro Tier, your content is not used for service improvement, or to train any underlying foundation models (FMs). Unless explicitly opted out, content from Amazon Q Developer Free Tier might also be used to enhance and improve the quality of FMs. Your content will not be used if you use the opt-out mechanism described in the documentation. For more information, see Sharing your data with AWS.”