What are the challenges for AI dubbing in audiovisual localisation?

AI is transforming audiovisual localisation through voice cloning, lip-syncing, script translation... But what are the risks for professionals, and what is the legal framework in Europe?

Introduction

Audiovisual localisation is no longer merely an “extra” for distributing a film or series beyond its domestic market. It has become a strategic infrastructure: without linguistic and cultural adaptation, a production faces rapid constraints in terms of audience, understanding and emotional resonance.

At the same time, the audiovisual industry is experiencing a rapid transformation. The growth of content libraries, expansion into multiple distribution territories and tight deadlines are pushing studios and platforms to industrialise their translation processes. It is in this context that automation —and especially generative AI— is being integrated into dubbing, offering the compelling potential to produce content more quickly, in multiple languages, and at controlled cost.

However, automation in dubbing is also sparking big debates.
Beyond the audiovisual sector, these debates reflect a broader transformation that today affects the entire translation and localisation industry, which is grappling with the rapid integration of artificial intelligence into its processes. According to Slator, in the article “German Voice Actors Challenge Netflix Over AI Dubbing - Slator”, German voice actors are concerned about Netflix’s increasing use of artificial intelligence. The German Voice Actors Association (VDS), representing around 600 artists, challenged a contractual clause granting a platform the right to use voice recordings to train AI systems. This tension illustrates a broader issue: how to innovate without undermining the rights, professions and artistic quality that give dubbing its value?

Why dubbing is essential in global distribution

Dubbing is not just a “spoken” translation. It is a performed adaptation: rhythm, intent, humour, cultural references, timing constraints... Everything must work in the target language while staying true to the original.

The rise of non-English content is one of the most visible drivers of localisation. In its engagement report for July to December 2025, Netflix states that non-English content accounts for more than one-third of total viewing.

The more these contents circulate, the more audiovisual localisation becomes a performance driver: high-quality dubbing can turn a “niche” series into a global phenomenon.

In France, Germany, Italy and Spain, a significant portion of the audience expects a “seamless” experience, without the need for continuous subtitle reading. Dubbing also addresses accessibility and inclusion issues (children, visual fatigue, mobile viewing contexts, etc.).

The arrival of AI in audiovisual localisation

Artificial intelligence is now integrated across multiple stages of the localisation chain, including script translation, voice creation and lip-syncing.

Voice cloning involves generating a synthetic voice based on samples from an actor or a reference voice.

The core issue is not so much feasibility as the conditions of use: who grants authorisation, for what scope, for how long and under what remuneration terms? In the German context, the key concern centred on the use of recordings for AI training, perceived as a potential path towards replacing human talent.

Some solutions use AI to adjust lip movements to match translated dialogue. Media reports also indicate that Netflix has tested this type of technology while continuing to rely on local voice actors.

Automation is also impacting upstream processes, including script translation, segmentation, named entity recognition, terminology consistency and glossary development. In audiovisual contexts, however, “raw” translation is almost never sufficient: it must be adapted, shortened, timed and maintain both intent and register. Human post-editing therefore remains a fundamental component when pursuing high-quality dubbing.

Risks for dubbing professionals

Innovation redistributes value. And in dubbing, this redistribution poses tangible risks: rights, remuneration, artistic quality and social acceptance. These concerns are also voiced by industry professionals. In recent industry discussions, voice actors worldwide have been vocal about the impact artificial intelligence may have on their profession, from AI voice cloning to automated dubbing workflows.

A performer’s recorded voice is not a freely exploitable resource: under the UK Copyright, Designs and Patents Act 1988, performers have rights over the use, reproduction, and distribution of their performances. When it is used not only for a work but also for training an AI system, it goes beyond standard audiovisual exploitation and enters the realm of fixation, reproduction and reuse of the performance, all of which require the artist’s authorisation.

If part of the market shifts towards “sufficient” solutions at a lower cost (low-stakes content, long catalogues, quick tests), pressure on rates may increase. Concerns focus on the loss of artistic value, standardisation and a drop in quality. These questions go beyond dubbing: they now affect the entire translation and localisation industry, where AI is gradually reshaping workflows, economic models and even the very definition of linguistic value.

Consent is not just a moral issue; it is a legal requirement, particularly when a voice is used as data to identify a person. In Europe, voice biometrics are subject to strict requirements, including clear purposes, data minimisation, security and individuals' rights).

In practice, this consent must be clear and traceable: authorisation for recording, model training and terms of use.

The legal perspective

While the term “voice data” is commonly used, it is important to distinguish between raw audio, metadata, voice templates or models and synthetic outputs. This distinction is critical for framing rights, ensuring GDPR compliance and structuring contractual clauses.

The European Union has adopted Regulation (EU) 2024/1689 (AI Act), which sets out harmonised rules for artificial intelligence, following a risk-based approach with associated obligations. For localisation teams, this calls for more robust governance frameworks, including traceability, documentation, quality control and clear policies on training data.

Transparency is key to rebuilding trust: with talent (actors, artistic directors), with clients (studios, platforms) and sometimes with audiences, depending on editorial decisions. A transparency policy built on clear, explicit and negotiated contractual clauses helps mitigate both reputational and legal risks.

Alternatives and solutions: towards a human + AI model

In audiovisual localisation, the most sustainable solution is often hybrid, because perceived quality and social acceptability matter as much as productivity gains.

A hybrid model may be appropriate when physical dubbing sessions are not feasible (due to urgency or logistical constraints) or when a performer is temporarily unavailable (e.g., loss of voice or short-term unavailability). AI then becomes an assistive tool, reducing friction, while human oversight ensures intent, performance quality and cultural adaptation.

Innovation is also creating new roles: AI QA specialists, audiovisual post-editing linguists and AI compliance managers.

Conclusion: what does the future hold for audiovisual localisation?

The debate around artificial intelligence in dubbing extends beyond a simple clash between innovation and tradition. It highlights a delicate balance between scalability, quality and respect for the rights of professionals. As noted by the specialised media outlet Slator, current discussions around AI dubbing show that the issue is as much technological as it is artistic and legal.

In the short and medium term, the most likely future is a hybrid approach to audiovisual localisation: AI can accelerate certain stages, but human expertise remains essential to ensure language quality, cultural adaptation and the credibility of content with audiences. This evolution reflects the ongoing transformation of the translation industry, in which AI is emerging as a powerful optimisation tool while not displacing the strategic role of language professionals.

In this context, a translation company plays a key role in supporting producers and platforms in managing localisation flows and quality control. For companies seeking to expand their international reach, working with an experienced translation company ensures the consistent adaptation of audiovisual and digital content, whether for dubbing, subtitling or website translation. With AbroadLink Translations, benefit from our expertise in audiovisual and digital localisation to effectively reach an international audience while ensuring quality and cultural coherence.