Hi RMatlack,
Short answer: we'd love to pull from org assets and are working on making that possible as we speak
Long answer: The reason we aren't currently pulling images from stock and not the 'your organization' folder is because stock images are keyword tagged. So, for example, an image of a dog holding a ball might be tagged with pet, dog, animal, ball, etc. This means that we can match keywords from your text to images in our stock library. Your org assets aren't keyword tagged, so we don't know what any of those images contain. Now there are 2 ways to solve this, 1. automatically keyword identifying org asset photos using AI, or 2. giving admins an interface to manually add keyword tags to images. They have their pros and cons: Solution 1 would be automatic but might feel invasive, whereas 2 would be tedious but extremely predictable.
Any thoughts whether either of these would meet your needs?