Protecting a single image is far more difficult. Video and audio are "protectable" to the extent they are because there is a really low expectation of how they have to work. Start playback, stop playback - that's about it. No one expects to pre-render, use video as a buffer for other things, etc. You can realistically render video direct to the screen via hardware, in a protected path.

Doing the same for still images adds a ton of work for little payoff. And it still doesn't prevent anything. At least in video if the system remains uncracked, it's hard to make a copy. You can't just point a video camera at a screen and get good results.

Whereas for a still image? Pretty easy to reconstruct the exact pixels, 100%.

Anyways, this is doomed entirely. They provide the case of social media sites stripping off metadata. Well guess what, if your DRM is gonna prevent them from modifying the image, it's gonna prevent them from loading the image in the first place. Even if the JPEG folks add it, it has zero chance of going anywhere. You'd need support from the hardware, OS, and on up. (Like video has.)