Apple is pushing hard to turn Fitness+ from an Apple Watch novelty into a genuinely global service. On December 15, the company will flip the switch on the biggest international expansion the product has seen since it launched: Fitness+ will arrive in 28 new markets — everything from India and Hong Kong to the Netherlands, Singapore and Taiwan — with Japan following on a slightly slower timetable early next year. Alongside the geographic rollout, Apple is shipping a technical change that matters almost as much as the new maps on the availability page: hundreds of workouts and meditations will now be available with digitally generated, trainer-matched voice tracks in Spanish and German at launch, and Japanese when Fitness+ goes live in Japan.

That last detail is what separates this from the dozen other streaming services that have added subtitles or hired local voice actors. Apple’s approach, at least as it describes it, is to generate dubbed tracks that are modeled on the actual voices of its Fitness+ trainers — the same cadence, the same pep, but in another language. Users will be able to switch audio tracks during a workout or pick a preferred language in the Fitness app, so dubbed audio plays automatically when available. For millions of people who’ve had to read captions while they sprint, that’s a genuine usability shift: following a HIIT session while squinting at your screen is no longer the only option.

Related /

• Apple Fitness+ is ‘under review’ as new bosses are set to decide its fate

There’s precedent for this in Apple’s recent AI work. Earlier this year, Apple described systems that build dynamic generative voices using voice data from Fitness+ trainers as part of its AI efforts — the company has been quietly stitching together the tech plumbing needed to do this at scale. That background helps explain how Apple can roll out dubbed episodes without re-shooting workouts in multiple languages: audio localization, not video reshoots, becomes the lever. It’s faster and far cheaper — and it preserves the on-camera trainer, which matters to a service built around personalities as much as exercise routines.

That efficiency is exactly why media companies and platforms have been experimenting with “AI dubbing” for a while. Business and localization teams love the math: localize audio rather than filming again, and you suddenly unlock new markets without multiplying production budgets. But technologists and linguists warn that the same tooling needs careful human oversight — literal translations can miss cultural beats, timing and coach cues must match the visual action, and music licensing needs get messier when playlists are globalized. An industry guide to AI dubbing notes those tradeoffs: speed and scale are real, but the best results blend automation with human review.

Apple is coupling the tech play with a couple of surface changes that are also strategically smart. A dedicated K-Pop genre is arriving on Fitness+ — simple, obvious and effective: K-Pop’s global audience is enormous and audiophiles already live inside Apple Music, so plugging those tracks into workouts is a low-friction way to boost engagement. Apple is also refreshing Time to Walk with a new episode from Formula 1 driver Yuki Tsunoda, a clear nod toward Asian markets as part of that broader push. Those are the kind of small product moves that make an offering feel local without the company trying to reinvent its entire editorial approach.

From a business perspective, the expansion matters because Fitness+ doesn’t live in a vacuum — it’s part of the services stack that hooks users into Apple Watch and iPhone hardware. More regions potentially mean more subscribers, and subscribers feed Apple Music, Apple One bundles and the broader ecosystem. Apple is even sweetening the deal with three months free for new device buyers in participating regions, the same tactic it’s used before to convert hardware purchasers into recurring-revenue customers.

But the very thing that makes AI dubbing useful — the ability to recreate a person’s voice in another language — raises thorny legal and ethical questions. Voice cloning and synthetic audio have become mainstream tools for media localization, accessibility and entertainment, yet they also carry real risks: unauthorized impersonation, deepfakes, and disputes over who owns a vocal likeness. Legal scholars and industry lawyers have been wrestling with these questions for months, and commentators warn that consent, transparency and auditable controls will be the backbone of any responsible rollout. The tech can work brilliantly, but without clear guardrails, it can also open platforms to reputational and legal headaches.

Those risks are not theoretical in every market. India, where Fitness+ launches December 15, has already seen high-profile legal action around AI voice and image cloning, underscoring how sensitive courts and regulators are becoming to misuse. That means Apple’s local teams will need to marry global product engineering with local legal and cultural awareness — and likely keep a close eye on consent flows and takedown mechanics for synthetic audio.

Practically speaking, for users, the update is straightforward: Fitness+ sessions will keep their original English audio but offer dubbed alternatives in Spanish and German from December 15, with Japanese promised when Japan goes live. You’ll need relatively recent Apple software — Apple notes the dubbing requires iOS 26.1, iPadOS 26.1 or tvOS 26.1 — and the usual device compatibility rules still apply (Fitness+ needs a modern iPhone or Apple Watch for the full synced experience). For people who travel or have family across borders, that means your morning run can now sound native, whether you’re in New York or New Delhi.

For the trainers themselves, the arrangement is worth watching. Fans pick instructors for their energy as much as for expertise, and Nike-level personalities can be a competitive moat for a fitness platform. If Apple has genuinely trained AI models on each instructor’s voice with their consent, it keeps the personality while letting the company scale. But the market is rapidly debating what “consent” means in this context — is it a one-time sign-off, a rights transfer, or an ongoing revenue share? Those are commercial questions the industry hasn’t fully standardized yet.

At the user level, the expansion is a clear win for accessibility and inclusion. Subtitles helped early adopters, but voice narration is a different sensory experience — especially for older users, people with visual impairments, or anyone juggling earbuds and a treadmill. If the generated voices really capture the coach’s timing and encouragement, they can make workouts more approachable and easier to follow. The real test will be in the details: naturalness of delivery, synchronization with music and on-screen cues, and how quickly Apple adds more languages beyond the initial three.

Apple’s Fitness+ becoming more multilingual is a small, concrete example of a larger trend: tech companies are applying AI to remove geographic and linguistic friction in services that were once expensive to scale. That’s good for consumers — more localized content with less delay — but it forces a new kind of responsibility on platforms to manage consent, provenance and potential misuse. For now, Apple’s playbook looks pragmatic: use the company’s technical muscle to expand reach, lean into music and personalities that travel well, and promise human oversight where the AI needs it. The rollout on December 15 will be the first public test of whether that balance holds.