You know the feeling: a character delivers a devastating confession in a dubbed foreign series, the English audio wraps up cleanly, and then their mouth keeps going for another half-second, like someone forgot to hang up the phone. Amazon has decided the most logical fix is to edit the actor’s face.
Starting with Maxton Hall, a German-language romance series that became Prime Video‘s most-watched international original, the platform now uses AI to adjust actors’ lip movements so they match the English dub rather than the original performance.
How It Actually Works
Amazon’s new system rewrites the visual side of dubbing, leaving the audio entirely in human hands.
The feature targets the video, not the audio. Human voice actors still record the English-language track the traditional way, with no synthetic voices involved, according to The Verge and Variety.
Amazon describes the result as “visual dubbing“: mouth movements in the footage are retimed or reshaped to accompany the translated audio. The image chases the words instead of the other way around.
Exactly how automated that process is remains unclear. Amazon says the feature was built using “a combination of AI and VFX tools” but has not disclosed the specific models, the degree of human intervention per frame, or whether the pipeline relies on proprietary technology or third-party solutions.
Raf Soltanovich, Vice President of Technology at Prime Video and Amazon MGM Studios, described the goal to Variety. “With lip-syncing, we’re creating a more seamless and immersive way for customers to enjoy global content,” he said, “all done under creative oversight to ensure the integrity of their artistic vision is preserved.” What that creative oversight looks like in practice, Amazon hasn’t said.
The Test Case and What Comes Next
Maxton Hall is Amazon’s highest-profile experiment yet in AI-assisted localization, and it won’t be the last.
The lip-synced English dub for Maxton Hall seasons 1 and 2 is available globally now. Season 3 arrives December 9, 2026, with the same treatment built in from launch.
In some regions, Prime Video labels the option explicitly as “Maxton Hall (with Lip Sync for English Dub),” separating it from the standard dub and subtitle choices.
This is a meaningful escalation from Amazon’s earlier experiments. Between 2025 and 2026, Prime Video tested AI-assisted dubbing on 12 movies and series, including El Cid: La Leyenda and Mi Mamá Lora, but those tests focused on audio localization only.
Maxton Hall marks the first time Amazon has publicly modified original footage at scale for a flagship scripted series. More titles are coming, according to Prime Video, though no names, languages, or timeline were provided.
Enthusiast communities on ResetEra have already flagged the obvious concern: will AI-adjusted faces drift into the same uncanny valley that made AI-generated hands so memorably wrong circa 2022? Those same communities are asking whether the technology has applications no press release would volunteer.
Everyone Else Is Already Here
Amazon’s approach differs from its rivals in one key way: it keeps real voice actors in the booth.
Meta automatically dubs and lip-syncs Reels for English-Spanish translation, according to PCMag. YouTube offers AI dubbing for creators across multiple languages. Both platforms lean on automated voice generation.
Amazon’s distinction is that it applies visual lip-sync to a professionally produced scripted series while keeping human voice actors in the recording booth, preserving the craft even as it adjusts the face delivering it.
The real question isn’t technical. If the sync is seamless enough that viewers cannot detect the edit, genuine questions about performer consent and the boundary between post-production and alteration of an actor’s likeness begin to surface. Those conversations are just getting started. The labeled version is live on Prime Video now, and the results are there for you to judge.
