There was already a way to make stock video, it's called "record some video". No reason to also hand some overfunded police commissioner the ability to frame any black person he likes of whatever crime he feels like or produce porn of random women on Facebook.
There is never any reason to adopt any tool based on the worse use case. We should have banned stones and wood because they can be thrown at people and made into a pointy stick and kept living in our caves if we had collectively decided based on the worst situation. But that's not how we do it generally. Adoption or outlawing of a technology depends on the cost vs benefits of it, and, as far as I know, housing is widely accepted today.
Sure, among the drawback are "someone might be framed by a made up footage that would escape scrutiny by the best expert available" and "someone might suffer from a deepfake being released". But, like with any tech, we don't focus only on the thousands of people dying in car crashes every year, we're focussing on balance between these casualities and the ability to drive to work and move goods at increased speed compared to a horse wagon.
Having anyone with limited resource (a computer) being able to create a stunning artwork on the level of the best films Hollywood produced, without needing to accessions billions in budget will enable everyone on Earth to benefit from free art. A small increase in utility for 10 billions humans can outweight a big decrease for a single one (car drivers vs car casualty). Having everyone on earth understand that they can't trust what they see on a screen will end propaganda that is currently available only to state actors, ending the tremendous power they have over masses. The resulting increase in critical thinking will usher a golden age of philosophy on the scale the printing press did. Videogeneration combined with VR headset will make the holodeck come true, creating a whole new artform and providing inexpensive entertainment for billions of people. It will allow people to live happier sex lives involving videotaping themselves without fearing that they leak -- everyone will think it's AI anyway. Virtual teaching assistant, combining LLM and voice generation, will be able to chat with kids, helping them in learning foreign languages, making them better in school and more educated citizens. There is a chance that sex crimes decrease if incels can spend their lives in their basement looking at porn instead of having to go out and kill women. At the very least, it will allow porn to be created without having porn actors being exploited. It will allow you to send an AI avatar attend a useless Teams meeting, intervening based on a prompt you fed it and summarizing the useless meeting outcome for you, saving tens of hours of precious, irreplacable lifetime for each and every human currently doing a naughty word job [*], freeing time for production of actual wealth and enabling societies wealth-redistribution mechanics to create new rights for the masses. The technological progress made on video-generating AI will also affect other field of AI, cascading into other use-cases with their own benefits.
Then, when the dust has settled, we'll collectively weigh the good and the bad and decide if (a) we allow it, like we allowed the Internet despite all the evil it brought over us or we're allowing selfies despite hundreds of people falling to their death making pictures, (b) we regulate it (like we allowed car, but require a driving license, or we allow the eating of raw food despite the health risk but we enforce sanitary control) or (c) we ban it, like jaywalking, because the benefits can't ever outweigh the risk. Not every society will make the same choice over it and that's good. Also, such determination can happen over a long time. The printing press was banned at some point in several countries, and pre-printing censorship was enforced for long before the benefit of free speech appeared to all.
Edit: * no, it's not a naughty word job. It's the kind of job aptly named by David Graeber in his eponymous publication.