> ## Content Index
> Fetch the complete content index at: https://www.thecriticalloop.com/llms.txt
> Use this file to discover other available public pages before exploring further.

# WHAT ARE THE PROSPECTS FOR VOICE ACTING WHEN AI CAN CLONE ANY HUMAN VOICE?
- URL: https://www.thecriticalloop.com/what-are-the-prospects-for-voice-acting-when-ai-can-synthesise-any-human-voice/
- Published: 2026-08-25T18:04:00.000Z
- Updated: 2026-09-02T09:28:08.000Z
- Description: Marco Nepomuceno knows that a memorable dubbed performance is built from decisions that never appear in a transcript. Drawing on his work as the Brazilian voice of the Merchant in "Resident Evil: Requiem 4", he considers what voice synthesis can reproduce, and what it cannot understand.
- Author: Marco Nepomuceno
- Tags: Marco Nepomuceno, Legislation-governance, Culture

By Marco Nepomuceno. 

In recent years, there has been a great deal of discussion about the use of artificial intelligence systems: their applications, their impacts and the challenges surrounding their implementation across a wide range of sectors. That includes Voice Acting. 

One of the most important elements in shaping this debate is the establishment of legislation that covers every aspect of the development and use of these technologies, as we have seen in various parts of the world, particularly in Europe. The impacts are wide-ranging, extending from the protection of intellectual property and employment rights to the management of energy and resource consumption. 

There are several high-profile cases that have demonstrated why a performer’s voice should be treated as an aspect of their identity requiring explicit protection, even though a voice itself cannot simply be copyrighted in many jurisdictions. 

Before retiring, the American Actor, [James Earl Jones](https://en.wikipedia.org/wiki/James%5FEarl%5FJones?ref=thecriticalloop.com) approved [Lucasfilm](https://www.lucasfilm.com/?ref=thecriticalloop.com)'s use of his archived recordings and AI technology to recreate Darth Vader's dialogue , demonstrating how such tools can be employed with the performer’s knowledge and consent. He also advised the team responsible for the digitally recreated performance in Obi-Wan Kenobi. By contrast, [Scarlett Johansson](https://en.wikipedia.org/wiki/Scarlett%5FJohansson?ref=thecriticalloop.com) took legal action after OpenAI released a ChatGPT voice that she said sounded remarkably similar to her own, despite her having declined an offer to provide the voice. OpenAI denied that it was an imitation but subsequently withdrew it. 

These contrasting cases illustrate the importance of informed consent, clearly defined limits on use, appropriate remuneration and the performer’s continuing control over any digital replica of their voice. They also show that protecting voice actors cannot depend on fame or on an individual artist’s ability to pursue legal action after the technology has already been deployed. 

In Brazil, there are several professional associations representing the artistic sectors and working alongside similar international organisations. These organisations have contributed to public debate and engaged with government bodies through a well-mobilised structure involving a number of voice professionals, and operating within a wide and well-established production chain. In 2025, the Observer reported on [how Brazilian Voice Actors have been fighting back against AI dubbing](https://observer.co.uk/news/international/article/brazils-beloved-voice-actors-face-ai-threat-to-dubbing?ref=thecriticalloop.com). Despite these efforts, the development and implementation of relevant legislation are slow and constantly at odds with the Big Tech lobby machine.  
  
Under Brazilian Legislation, voice acting is already recognised as a distinct professional activity, and the country has developed a substantial dubbing infrastructure over several decades. Its reputation rests not on an official international ranking, but on the maturity of the sector, the training and dramatic ability of its performers, and an approach to localisation that adapts humour, idioms, rhythm and cultural references rather than merely translating dialogue word for word. 

The continuity between actors and characters across films, television series, animation and games has also allowed Brazilian voices to become part of audiences’ emotional connection with international works. This contribution has occasionally been recognised by the original performers themselves. [Hugh Jackman,](https://nypost.com/2022/02/02/hugh-jackman-honors-late-wolverine-voice-actor-isaac-bardavid/?ref=thecriticalloop.com) for example, publicly acknowledged [Isaac Bardavid](https://en.wikipedia.org/wiki/Isaac%5FBardavid?ref=thecriticalloop.com), his long-standing Brazilian voice actor for Wolverine, as sharing in the affection that Brazilian audiences had developed for the character. 

Dubbing is particularly important in Brazil. It is not a marginal alternative to subtitling, but the principal means through which much of the population encounters foreign audiovisual content. A [2008 Datafolha survey ](https://www1.folha.uol.com.br/fsp/ilustrad/fq2908200808.htm?ref=thecriticalloop.com)of cinema audiences found that 56 per cent of Brazilians preferred dubbed films, compared with 37 per cent who preferred subtitled versions.  
  
Industry data for 2014 subsequently showed that dubbed productions accounted for 59 per cent of Brazilian cinema admissions and 57 per cent of box-office revenue, while subtitled films accounted for 28 per cent of admissions. Although these figures predate the expansion of streaming, they demonstrate the established scale of the preference that streaming services later inherited.

Dubbing increases the accessibility of foreign movies. It opens access for children, people with limited literacy, those who find it difficult to follow fast-moving subtitles and viewers who do not understand the original language. This is especially significant in a country in which English proficiency remains uneven. 

The 2025 [EF English Proficiency Index](https://www.ef.co.uk/epi/?ref=thecriticalloop.com) ranked Brazil 75th out of 123 countries and territories, placing it within its “low proficiency” category. Dubbing is therefore not simply a matter of consumer preference; it is also a form of linguistic and cultural access. The growth of digital gaming has further expanded the importance of Brazilian voice acting. 

The [2026 Pesquisa Game Brasil,](https://www.pesquisagamebrasil.com.br/?ref=thecriticalloop.com) based on 7,115 respondents aged between 16 and 55, found that 75.3 per cent of Brazilians consume digital games, while 86.7 per cent regard games as one of their principal forms of entertainment. Brazilian Portuguese was already the fourth most requested language for game localisation among developers surveyed by [Localize Direct ](https://www.localizedirect.com/?ref=thecriticalloop.com)in 2020, accounting for 8.5 per cent of demand. 

For narrative-led games, localisation can involve thousands of lines, numerous characters and performances that must respond to gameplay rather than to a fixed sequence of scenes. It is consequently both a commercial requirement and a substantial field of artistic employment. This also makes the replacement of performers with synthetic voices particularly consequential. 

In the [2026 Pesquisa Game Brasil](https://www.pesquisagamebrasil.com.br/?ref=thecriticalloop.com), 45.7 per cent of respondents expressed concern about job losses and the weakening of creative work through AI, 39.6 per cent were concerned about the unauthorised use of artists’ work, and 38.4 per cent feared that AI-generated games would suffer a decline in quality.

There are technical areas in which tools employing these technologies, when operated by professionals experienced in the sector’s methods and standards, can make processes more efficient. It has become common to regard these benefits positively. However, the first AI-assisted productions released by streaming platforms, including films and documentaries, were not well received. 

From an artistic perspective, the impact was extremely negative, prompting severe criticism from audiences. The expression ‘It looks as though it was made by AI!’ has already become a common way of describing audiovisual content of questionable quality. Some of these productions have even been withdrawn.

Where the artistic dimension of a product is less important, and its purpose is purely commercial or informational, and it is intended for immediate publication, the adoption of such methods is understandable. The same applies to supporting materials and functional accessibility resources. Despite this, developing alternatives that retain the direct involvement of human professionals demonstrates not only care for the work itself, but, above all, respect for the people who will consume the content.

One possible way forward, worthy of analysis and discussion, would be to place these tools directly in the hands of artists, enabling them to develop their own work, rather than leaving them under the control of large corporations that rely on mass production, misappropriation and one-sided agreements.

Another consideration is the risk of cultural, symbolic and artistic homogenisation, through which features of local cultures may be erased as part of the exercise of *soft power*. It is well known that current implementations of these technologies are saturated with biases that compromise their usability. 

When carried out properly, localisation involves analysing the original work and reinterpreting it for the local culture without erasing its original context or intent. Subjecting these choices to statistical assessment may lead to a significant loss of the sensitive and emotional dimensions characteristic of each people, as well as their linguistic and expressive forms and symbolic points of reference.

Dubbing demonstrates the enormous difference between hearing an audiovisual work in your language and having it performed in your language. Although the two descriptions may sound similar, they are far from equivalent. Dubbing involves more than translating phonemes. It requires adapting the text and conveying an appropriate dramatic intensity that corresponds with the scenes and the way they are received.

At times, the statistically optimal choices, those closest to 100 per cent accuracy, are not the ones that best serve the work’s purpose. When sensitivity and aesthetic quality are involved, it is often the unpredictable, immeasurable and spontaneous element of immediate human performance that enables audiences to experience a product as less global and more local. Added to this is the collective nature of the process, in which everyone involved can contribute their life experience and their own universe of meaning. Aware of this aspect, many creators of audiovisual works, particularly in the digital games industry, have publicly declared their commitment to preserving the human artistic component of their work, as they understand it cannot be replaced.

The use of technology in art, and of electronic equipment in its creation, is nothing new. Art and technology have, in fact, always been closely connected. Human perception, however, can be adjusted, trained and regulated. To someone who has very little, even a small addition may feel like a gain. Take everything away, wait for a while and then offer the bare minimum, and an entirely new perspective emerges.

We may be moving towards a point in human history at which our sensibilities become accustomed to works created by artificial intelligence tools. Perhaps, in the process, something within our capacity to feel will be transformed, or lost.  
  
[Resident Evil 4 Remake - #02 -The Merchant -Voice Over in Portuguese ](https://youtu.be/1uGbCCO7dMk?si=jyXhDp%5FxQV4vHBiU&t=405&ref=thecriticalloop.com)