As Disney and Universal anticipated in their Complaint against Midjourney, a new feature of video generation has just been launched by Midjourney.
And the results are super impressive:
but is this risky business for midjourney?
One of Disney and Universal’s main gripes is that Midjourney allegedly failed to institute greater guardrails to prevent Midjourney from generating Disney and Universal characters, even after their lawyers sent Midjourney cease-and-desist letters.
One of the reasons the movie studios sued, I gather, is that they were concerned with the upcoming use of their characters in videos as opposed to just images.
Well, I tested Midjourney Video out. And the guardrails either haven’t changed or are not that effective. I created this video of Hiccup and Toothless. (Please don’t sue me, Disney and Universal. This is fair use–and just a test of the allegations in the Complaint.) The video is very impressive.
I created another one with Homer Simpson:
Here’s one with Shrek skateboarding:
However, I wasn’t able to create a video with Darth Vader, Spiderman, Olaf–and received a message “Creation failed.” Midjourney does create images of those characters. But I wasn’t able to generate a video. It could be an example of some video guardrails on Midjourney, but I will have to keep testing to eliminate the possibility of some other reason or glitch.
UPDATE: I suspect Midjourney does have some guardrails in place, although I am not 100 percent sure. It didn’t create videos of Spiderman, Superman, Darth Vader, Yoda, and Olaf when I prompted it. But it did create a video of Darth Vader and Luke Skywalker. Unlike the videos of animated characters above, this video is less impressive because it doesn’t look exactly like the real people in the movie. What do you think?
The motion for preliminary injunction hearing just got more interesting.
UPDATE on June 21, 2025 morning:
It looks like Midjourney has increased the guardrails to stop the creation of Disney and Universal characters in videos, while allowing them in images.
I got this message after trying to generate a video of Toothless and Hiccup. The message states: “Sorry! The AI Moderator is unsure about this prompt. Reasoning: This may be a violation of Midjourney’s terms of Service. Please try another prompt.”
For whatever reason, this message didn’t appear yesterday morning when I created the sample videos above. But it does today.
Drawing a line between images and videos in terms of guardrails will no doubt raise questions in the lawsuit. If the guardrails can stop videos of copyrighted characters, why aren’t they being implemented to stop images of them, too? I’ve already mentioned, though, that I think Disney and Universal were either most or especially concerned with the video generator they anticipated in their complaint–which may explain the timing of its filing. Nonetheless, if the guardrails are effective in stopping generation of copyrighted characters in videos, Disney and Universal likely will use that as evidence that comparable guardrails can do the same for images.
Related Stories