Yes, if it’s truly metadata that’s not in the image itself. For instance, it could theoretically be digitally watermarked (this technology already exists, actually) in a manner that humans can’t see or is tough to notice, but an algorithm looking for it can spot. That can be defeated, too, although depending on the robustness of the watermark technology it may take more effort.
The output loophole always exists: Any time you produce any output capable of being understood by a human (eyes, ears, both…) somebody can record and reproduce it. Probably not bit-for-bit, pixel-for-pixel, but you can always point a camera at the screen. (Or put your screen face down on a flatbed scanner that’s had its lightbar defeated, or put a microphone in front of the speakers, or…)
Yes, if it’s truly metadata that’s not in the image itself. For instance, it could theoretically be digitally watermarked (this technology already exists, actually) in a manner that humans can’t see or is tough to notice, but an algorithm looking for it can spot. That can be defeated, too, although depending on the robustness of the watermark technology it may take more effort.
The output loophole always exists: Any time you produce any output capable of being understood by a human (eyes, ears, both…) somebody can record and reproduce it. Probably not bit-for-bit, pixel-for-pixel, but you can always point a camera at the screen. (Or put your screen face down on a flatbed scanner that’s had its lightbar defeated, or put a microphone in front of the speakers, or…)