“Adobe does not train Firefly Gen AI models on customer content. Firefly generative AI models are trained on a dataset of licensed content, such as Adobe Stock, and public domain content where copyright has expired.”

This references a single particular product. lol. If they're training a model by a different name with customer data, it would still be a true statement.

The points about lawyers and NDA's hit the nail on the head. I thought something similar with the Windows Recall debacle. That's a juicy set of data for anyone looking to find journalist sources or scrape a hospital's network. In every case it relies on the end user (business or individual) to know how to disable those features with GPOs/registry options... There's no way 100% of them realize the issue and have the knowledge to fix it.