There’s a specific kind of discomfort that comes from hearing something that sounds human, almost gets it right, and then misses the mark just enough to remind you that it isn’t. Over the past few months, that feeling has crept into mainstream media, and not just as horrible dubs. Artificial intelligence has begun speaking, not metaphorically, but literally, on behalf of major companies, like Amazon and Apple. The result hasn’t been outrage so much as unease, a collective pause if you will.
Amazon’s experiment with AI-generated dubbing on Prime Video was one of the first moments where that discomfort became impossible to ignore. Late last year, as first reported by Futurism and IGN, the platform introduced AI-generated English and Latin American Spanish dubs for select foreign films and anime series. They initially labeled them as “AI beta.” On paper, it sounded reasonable that AI could help localize older or niche titles that might never receive a traditional dub without the cost of full production. Unfortunately, the experience of actually hearing those dubs told a completely different story.
In clips that circulated on X, the characters spoke clearly and fluently but lacked emotional grounding. In one scene meant to play as a tearful deathbed confession, the AI voice delivered its lines in a casual, upbeat tone, like a weather reporter predicting mild rain. The words and timing were technically correct, yet the scene felt fake. For anime fans, who are used to voice acting performances, the effect was lost. Prominent voice actors, including “Banana Fish” alum Daman Mills, condemned the move on social media, calling it “disrespectful” and “erasure” of the art form.
Within weeks, Amazon removed several of the AI-generated English dubs from its platform, including “Banana Fish” and “No Game, No Life.” There was no press release, no walk-back, just a silent reversal that said more than any statement could. Viewers pushed back, of course, loudly and collectively, and for a moment, it seemed they had won.
But the company’s ambition didn’t vanish with the dubs. In late January, after the initial outcry, outlet Anime Corner spotted a job listing for a “Global AI Dubbing Creative Director.” The role, which GamesRadar later verified, implied that AI was not just a failed experiment but a permanent strategy intended to drive creative localization at scale. The listing went viral and disappeared within a day, but the signal was clear: they aren’t stopping, they are just reloading.
Apple’s entry into the AI voice conversation unfolded more smoothly but left a similar sense of unease. In December 2025, the company announced a massive expansion of Apple Fitness+, bringing the service to 28 new international markets. To make that scale possible, Apple used AI-generated voices to dub workouts and guided meditations into Spanish, German, and Japanese, with the voices modeled after real Fitness+ trainers.
For many users, something still felt off. Fitness content, as a whole, relies on motivation and that almost-real sense that the trainer is sweating just as much as you are. You are not just following instructions; you are being encouraged by someone who sounds invested in your effort. It’s because of this that critics questioned why a company with Apple’s cash reserves would automate motivation rather than hiring local trainers.
Apple’s smoother rollout is not a technological failure, but economic logic. As entertainment analysts have long noted, streaming platforms are no longer in the movie business. Success is measured in volume, reach, and constant output. When a service needs to produce thousands of hours of content across dozens of languages, human performers seem to become more of a hurdle for algorithms to remove.
To the leaders of these projects, AI voices offer speed, scale, and cost savings. They don’t demand schedules, contracts, or residuals. From a corporate standpoint, the appeal is obvious. From a cultural standpoint, though, the cost is way harder to quantify. Audiences are not rejecting AI outright, but rather what it replaces. It falters when it attempts to stand in for emotion, intention, and presence. A voice is not just sound but a form of trust. It’s clear resistance might be the only way since Amazon pulled the dubs after audience backlash.
AI voices are not going away, of course. If anything, they will become more refined and harder to detect. One thing is clear, when it comes to entertainment, sounding human isn’t the same as being human, and the audience knows the difference.
