Pollo AI has launched Pollo Recast, an AI video recreation tool that rebuilds an existing video around a user’s own character, product or setting while keeping the source clip’s motion, pacing and camera work recognisable. The company positions it against tools that change only one layer of a video — motion transfer, style transfer, face swap — arguing that Recast recreates an entire clip: the source video sets the performance, and the user’s images decide who and what appears. It is available now, with free credits to try.
The intended uses are trend-driven. A viral dance can take an original character, a mascot or a brand’s team; a cinematic reveal can take a new product; a meme format can localise for different markets with local people and settings. Pollo’s chief executive, Bill Zhu, framed the pitch around the idea that “great content rarely comes from a blank prompt,” and that Recast starts from a video that already works.
A category that already exists
The distinction Pollo draws is fair, but it describes a category rather than a first. AI video recreation tools broadly split into three kinds: face swap changes only a face; motion transfer gives a character image the movement from a reference clip; and whole-video recreation re-renders the entire clip, keeping the shots, cuts and timing while replacing who or what appears. Pollo Recast belongs to the third kind, and independent comparisons already place several tools there — Higgsfield’s Genjutsu and Recast, Riffkit’s Swap mode, Kling’s editing tools and MiniMax’s Recast among them.
So the accurate claim is that Recast competes in an established bucket, not that it invented one. What it adds is a familiar packaging choice: a template gallery, a one-click recreation flow and free credits, aimed at creators and marketers rather than specialist video editors.
The question the release doesn’t answer
The more consequential omission is rights. A tool built to recreate a video the user “already loves” and recast it with their own characters and products implicates two things the release never mentions: rights in the source footage, and rights in any real person’s likeness used as a reference image.
That is not a peripheral concern, because the law on both is unsettled and moving. In the United States, the Take It Down Act became law in 2025, and proposals such as the NO FAKES Act seek a federal right of publicity covering AI replicas of a person’s voice or likeness, while many states have passed their own deepfake statutes. In the UK, there is no single personality right; protection rests on a patchwork of passing off, data protection, trade marks and defamation, and the government has consulted on whether to introduce something more specific.
Independent tool guides routinely warn users to replace a person’s likeness only with consent or the necessary rights, and to check each platform’s terms — advice that matters more, not less, as these tools get easier to use.
Pollo’s release does not state its policy on consent, on likeness, on watermarks, or on what happens if a user uploads footage they do not own. For a product whose core promise is remixing content that already exists, that silence is the story.
What to watch
Three things will show whether Recast is more than a template gallery. First, whether Pollo publishes the technical specifics — model, resolution, clip length and pricing — that the release omits. Second, whether it publishes a rights and consent policy, and how it compares with competitors that already warn about likeness. Third, whether the tool handles the practical failure modes the category suffers from, such as identity drift, where a character’s appearance changes across a clip. The announcement addressed none of these.
What the launch does and doesn’t establish
It establishes a new tool that recreates whole clips around custom characters, products and settings, a template-based workflow, and free credits to try it. It does not establish any technical specification, any pricing beyond the free tier, any named customer, any independent quality benchmark, or any position on source-footage and likeness rights.
The honest summary is that Pollo Recast is a competent entrant in an AI video recreation category that already has capable members, and its novelty lies more in packaging than in capability. The feature is easy to describe and the risks are easy to overlook: a tool that turns any video into a template for someone else’s characters and products raises consent and IP questions that the announcement does not begin to answer, and in AI video recreation, a field where courts and legislatures are still working out what the rules should be, that silence is worth noting as much as the launch itself.
Editor’s Note: This article draws on the Pollo AI release and its product page, independent AI-video tool comparisons, and legal commentary on deepfake, right-of-publicity and personality-rights law. The tool’s capabilities are as described by the company; no technical specifications, pricing detail or rights policy were disclosed. No interview was conducted.