Here’s an understatement: AI in motion design is evolving at lightning speed. As a motion graphics designer, I’ve been intrigued and a little overwhelmed by the explosion of creative AI tools in everyday workflows.

After Envato launched its own AI video generation tool, VideoGen, ignoring this shift wasn’t an option. So I dove in — asking around the community, testing new tools, and learning how AI tools for animators and motion designers are changing the game.

From prompt engineering for visuals to style transfer, here’s a practical look at how generative AI for creatives is being used by pros and the AI workflows for motion designers shaping the future of video content.

Image and video generation tools are probably the first thing you think of when you think of AI. Platforms like ImageGen can produce everything from photorealistic fashion shoots to the trippy images we see all over the internet today. These platforms use algorithms trained on large data sets (models) to produce images based on written or visual prompt inputs — it’s what gives you the ability to turn your profile picture into a Studio Ghibli-esque illustration or a dragon, as you prefer.

I have to be honest, I couldn’t initially see the fun in prompting as a form of creative expression; wrestling with AI to get the result I wanted felt tiresome. But I‘ve encountered a few main workflows that alleviate many of my former concerns. The first is training your own models.

Vitor Texeria, from the 3D-focused, Madrid-based studio NotReal, has been experimenting with creating his own data sets trained on the distinct style of NotReal.

He impressed on me the importance of owning the data that feeds the AI:

Vitor’s process includes using NotReal’s portfolio of past work to train specific models in an AI tool. He then uses these data sets to generate imagery that can become moodboards for art directors and 3D designers in the initial conceptual stage of a project.

This kind of workflow reflects the growing role of generative AI in 3D, where image creation and scene ideation are becoming more efficient and scalable with the help of AI tools.

Most motion designers I spoke to were using generative AI in this way—not as a means to an end but as a tool for quicker iteration in the early stages of a project, trialing a wider and more specific range of ideas and then taking them into their traditional pipelines.

I’ve also encountered some interesting workflows that use generative AI tools past the point of ideation. The music video clip for Cuco’s A Love Letter to LA used some pretty groundbreaking generative techniques to scale and speed up a traditional animation workflow.

ComfyUI is an open-source, node-based program that grants you a lot of control over parameters and feels similar to working in Blender or Houdini. Director Paul Trillo and his team of animators created complex node systems in ComfyUI that allowed them to input a cel-animated sketch and a filled style frame and output a fully painted cel animation.

Other motion designers and animators have used a similar workflow, where the AI makes itself useful in the tweening stages while human animators and illustrators handle the keyframes. 

Style transfer is another good use for AI. For A Love Letter to LA, illustrations by local artist Paul Flores were used to train a model that could then be applied at the end, like a filter. This approach can combine the varied styles within a team of animators into one unified style at the click of a button.

In terms of my own process, I’ve found generative AI useful in extending photo backgrounds and creating still assets for use in broader, collage-like animations (a method that’s increasingly common in emerging 3D design trends).

For instance, check out this statue. I couldn’t find a good image on a stock site, so I made it using AI. I was able to prompt details like the phone and the bird to better suit the project’s messaging. This approach has saved me lots of time searching for the perfect stock photo.

I have found that playing around with how you construct your prompt is very important. It’s akin to having different render engines. For example, I tested the above animation using AI to see what results I would get, and they were interesting.

Prompt:

Start frame:

VideoGen with prompt enhance:

Without prompt enhance:

For short, specific moments like this scene, where, for instance, animating the bird realistically would take a long time, I could see certain models being super helpful. But for the day-to-day precise keyframing work, I tend to agree with Vitor:

Aside from the image and video generation tools, there are, of course, many more use cases for other kinds of AI that are more closely related to a supercharged plugin.

For example, Cavalry motion graphics has emerged as a flexible platform for procedural animation, especially when combined with AI-generated elements in data-driven workflows.

A quick survey of industry professions reveals some common threads.

Whether it’s building custom animation presets or developing unique workflows, motion designers continue to experiment across tools—many drawing inspiration from artists like Ben Marriott, who regularly explore both technical and stylistic techniques in motion design.

For a deeper dive on AI coding, you can read our recent article Vibe Coding in After Effects.

Mask Prompter for After Effects uses AI technology within an After Effects plugin to quickly and accurately rotoscope your video.

Bullet Time allows you to up the frame rate of your footage using AI for super slo-mo effects.

If there’s one key takeaway from my conversations and experiments, it’s this: to create anything truly worthwhile with AI, you still need vision. The output quality depends entirely on the effort, creativity, and ideas you bring to it.

AI is a powerful tool for iterating on ideas. Motion graphics is a precise and time-consuming art, and AI can help speed up some of the more tedious tasks, like rotoscoping or sourcing stock footage. It can also open up new possibilities and scale projects to levels that were previously out of reach, as in the case of A Love Letter to LA.

Interested in a little further reading? Check out our complete guide to creative AI video prompts. Secondly, try this great talk from Joey Koreman on the future of Motion Design and AI, look at this crazy 3D AI model generation, or learn how to create your own 3D textures from photos.