Prime Video starts using AI to match actors' lips to dubbed dialogue

Prime Video's AI lip-sync arrives

According to The Verge, Amazon's Prime Video has begun rolling out a feature that uses AI to align an actor's mouth movements with dubbed dialogue. The notable detail is that the dubbing itself is still performed by humans: the AI steps in afterwards to reshape what viewers see on screen, so the picture matches the translated speech rather than the original language.

The feature debuts with the English dub of Maxton Hall, a German series. As reported, it currently covers seasons 1 and 2 of the show and is available to subscribers worldwide. Prime Video says the lip-synced dub will also be in place for the show's third season, which arrives on December 9th, and the company intends to expand the technology to additional titles over time.

How the technology works

Per Prime Video's announcement, the system combines AI with visual effects technologies to adjust lip movements so they fit the translated audio track. The framing positions this as machine learning working alongside traditional VFX rather than fully synthetic dubbing: the voices remain those of human performers, while the image is manipulated to fit them.

That distinction is worth noting, because much of the recent activity in this space involves generating both the translated speech and the synced mouth movements automatically.

Part of a wider industry push

The Verge points out that Meta and YouTube have both recently launched AI-powered auto-dubbing tools for creators, including an optional lip-sync mode designed to make it look as though a person is genuinely speaking in the viewer's preferred language. Prime Video's approach differs in two ways: it is being applied to scripted, professionally produced television, and it is layered on top of conventional human dubbing rather than replacing it outright.

Nor is this Amazon's first step into AI-assisted translation. According to The Verge, last year Prime Video experimented with what it called AI-aided dubbing in English and Spanish across 12 movies and TV shows, including El Cid: La Leyenda, Mi Mamá Lora, and Long Lost. The Maxton Hall rollout represents a more visible evolution of that work, moving from the audio pipeline into visual manipulation of the finished picture.

Why it matters

Dubbed content has always carried a built-in credibility problem. When mouth movements clearly belong to one language and the audio to another, viewers are constantly reminded they are watching a translation, which is one reason subtitles retain a devoted audience even as streaming platforms push dubs for accessibility and reach.

For a global streamer, that friction has real commercial weight. Non-English originals such as Maxton Hall are among the most valuable assets a platform owns, and their performance in markets like the United States depends heavily on how natural the English dub feels. AI lip-sync attacks the single most visible giveaway of dubbing, and doing it on top of human voice performances lets Amazon claim it is augmenting dubbing workflows rather than automating voice actors out of them.

The move also signals where the technology is heading. If lip-sync correction proves reliable on a flagship series, it becomes easier to imagine it applied broadly across back catalogues and to lower-budget localisations that previously could not justify frame-by-frame VFX work. That raises questions the announcement does not address — among them how performers and productions feel about their likenesses being altered, and where the line sits between enhancement and manipulation. For now, the technology is opt-in from the platform's side and limited to one show, but a single-series debut is a familiar pattern for streaming features that later become defaults.