Google is preparing to put generative AI video creation directly inside YouTube Shorts, turning a tool that once looked experimental into a feature meant for everyday creators. The planned update will use the company’s Veo model and allow users to generate short video clips from text prompts, then post them as part of Shorts.
The move matters because it brings AI video generation into the same app where creators already edit, publish, and compete for attention. It also raises familiar questions about authenticity, creative control, spam, and how viewers will understand what they are watching.
What YouTube plans to add
The new tool is planned for 2025. Sarah Ali, a senior director of product management at YouTube, said users will be able to create stand-alone video clips and shorts, including six-second videos generated from an open text prompt.
The idea is not limited to background effects or decorative edits. YouTube is moving toward a feature that can create fresh video material for Shorts directly inside the app. That means a creator looking for extra footage, or trying to show a scene that would be hard to film, could ask the system to generate a short clip instead of searching for existing media or using a separate tool.
Google has already tested generative tools in the Shorts ecosystem. Over the summer, it launched Dream Screen, an experimental feature that creates AI backgrounds for videos. Before the broader clip-generation rollout, Google plans to update Dream Screen with Veo in the next few months.
That sequence shows a clear path: first AI-generated backdrops, then full AI-generated clips. For creators, the practical difference is significant. A background tool helps frame a video. A clip-generation tool can become part of the video itself.
Why Veo is the center of the plan
Google has shown several AI video models in recent years, including Imagen and Lumiere. For this YouTube integration, however, the company is consolidating around Veo.
Eli Collins, a vice president of product management at Google DeepMind, first demonstrated generative AI video tools to Google’s board of directors in 2022. At that point, the system was slow, costly to run, and sometimes produced odd results. Even so, the ability to create video from a random prompt made the technology feel important enough to pursue.
Now Google is trying to move beyond the demo stage. Collins described Veo as the model the company will focus on in the near future, rather than presenting a series of separate video models. The YouTube Shorts feature is a major test of that strategy because it puts the model into a real creator product, not just a research showcase.
That distinction is central to the rollout. A model can look impressive in a controlled demonstration, but YouTube Shorts requires something broader: a tool that is useful, scalable, and deployable for creators working inside a live product.
Creators get new options and new concerns
YouTube’s argument is that Veo can support creators rather than replace them. Ali has positioned the tool as a way to help people find footage for a video or imagine scenes that would otherwise be difficult to produce.
That could be useful for creators who need quick visual inserts, fantasy imagery, or short scenes that support a larger idea. A six-second AI video may not replace an entire creative process, but it can change the amount of production effort needed to complete a Short.
At the same time, generative AI remains contentious among creators. Some see the current wave of AI tools as taking from creative work and weakening the human process behind media. Those concerns do not disappear just because the tool is built into YouTube’s own app.
There is also the issue of sameness. AI art can have a recognizable look, and video creators who care about personal style may not want their work to feel as if it came from the same model as everyone else’s. Collins has said that avoiding an obvious DeepMind signature in Veo outputs is important, because the goal is to make the model adaptable to what a creator is trying to make.
Labels, watermarking, and the feed problem
YouTube says AI-generated Shorts will carry disclosure signals. Google plans to watermark every AI video generated for YouTube Shorts with SynthID, which embeds an imperceptible tag to help identify synthetic media. The platform also plans to include a “made with AI” disclaimer in the description.
Those measures are meant to help viewers understand how a video was made, especially when the production method matters to the audience. They do not, by themselves, answer every concern about volume or quality.
One obvious risk is that easier video generation could encourage more low-effort Shorts. The source article notes that hustle-culture influencers already use multiple third-party tools to automate parts of the creative process and try to make money with minimal effort. If YouTube lowers the barrier further inside its own app, more people may try to publish AI-made videos at scale.
YouTube’s answer is that its recommendation experience and standard guidelines still apply, regardless of which tool was used to make a video. In other words, AI generation does not create a separate rulebook. Shorts made with Veo will still enter the same environment where YouTube tries to match content with viewers and enforce platform rules.
A small clip with large implications
The planned Veo integration is not simply a new editing button. It is a sign that generative AI video is moving from specialized tools into mainstream creator workflows.
For YouTube, the bet is that creators will use AI clips as building blocks, not as a substitute for audience understanding or creative judgment. For Google DeepMind, the feature is a chance to bring an AI video model into a creator product at a scale it has not previously handled.
The central question is what happens after the novelty fades. If Veo helps creators make clearer, more imaginative Shorts, the tool could become part of ordinary production. If it mainly increases synthetic filler, YouTube’s recommendations, guidelines, labels, and creator incentives will be tested quickly.
Either way, YouTube Shorts is about to become a major proving ground for AI video. Six seconds may sound brief, but inside a fast-moving feed, that is enough time to change what creators make and what viewers expect to see.