By Alana Ballantyne

“Localization” is not a word that gets a ton of air time outside the entertainment industry. Not many people outside the industry have any idea what the word means, though the size of the localization sector is valued at $56.18 billion dollars. For those not in the know, “localization” is the term given to the process of editing films to appeal to a specific regional audience. This can include the dialogue from English to the local language, but may also include changing characters, removing scenes, or editing references to make a joke or visual gag understandable to an international audience. It’s far from glamorous, and in an industry that thrives on glamor, the localization of film receives little notice.

But it should. In a cultural moment where Hollywood is abuzz with growing concern about the impact of generative Artificial Intelligence (“AI”) on the entertainment business, this little corner of the business has been quietly overtaken – almost entirely – by high-quality AI technology. 

Historically, in order to generate a high-quality end localized entertainment, studios have had to invest heavily in local talent. When a show was produced in a market, it was then prepped for international distribution. First, local writers and translators were hired to comb through the script – translating dialogue, creating subtitles, and adapting references and jokes to the target market. Then, local voice actors needed to be found to dub the translated dialogue. Producers and attorneys had to be brought in to oversee the process and ensure that the translated content did not run afoul of local regulations. 

The process of localizing a product for a target market was incredibly intensive. And these studios weren’t just doing this for one country, but dozens upon dozens of markets the world over. According to a September 2023 report from “Variety” some of the biggest players in the entertainment sphere are spending upwards of $30 billion dollars annually on content localization. Disney alone spends $33 billion per year on localization, while Netflix clocks in around the same mark. 

These massive costs benefit the company in the end; if the content is localized sufficiently, it will appeal to international audiences and make the company money. 

Generative AI has changed the game. 

AI can take original footage and translate it at the press of a button, syncing dialogue very closely to how the actor’s mouth moves on screen. Significant visual lag times, when the actor stops speaking long before the dub, or vice versa, are swiftly becoming a thing of the past, thanks to generative AI. Not only that, but the tool can mimic an actor’s actual voice. This means that rather than a local actor being hired by a studio or production company to dub a film or television show, footage of – for example – Chris Pine – can be run through an AI tool and the result sounds like Chris Pine is speaking Chinese.

This process is significantly faster and cheaper than traditional dubbing methods. Cost reductions estimate that properties which use AI dubbing save between 30% to 40% on localization when all is said and done. 

This technology will only become more advanced as time progresses – automated localization has made huge gains in just the past three years – and it is already at the point where it is economically preferable to hiring entire local teams. Could this technology one day be used to generate off-camera dialogue – cutting down on an actor’s need to be on set?  Could it be used to gap-fill in editing, avoiding costly re-shoots? What about the more than 640,000 professional translators and interpreters currently contributing to the global language services market? The way this technology is trending, those jobs will soon be obsolete. 

The “machine translation market” – another term for generative AI dedicated to localization – is projected to reach $1.5 billion in localization market share by 2024. In a sector valued at over $50 billion dollars, AI certainly has a long way to go before it entirely captures the market. But when asked in a 2020 survey about the trend with the largest potential impact on the industry over forty percent of Language Service Providers (“LSPs”) indicated that AI integration is poised to change their industry. 

That is a bet that many emerging companies are willing to take. 

Deepdub, an Israel company looking to carve out a place in the localization market by focusing their AI tool on older content which was not localized for certain markets because studios did not see a way to turn a profit. Perhaps the language was too niche; perhaps it was impossible to find a reliable translator, or perhaps a translator could be sourced, but the investment was not worth the forecasted payoff. Now, with Deepdub’s AI tool, these films and movies can be localized for a variety of markets previously shut out due to market conditions. 

“Localization is the biggest bucket that we’re working on for film and TV,” Ofir Krakowski, the CEO and co-founder of Deepdub, told Variety last year. “It’s mostly library content that was not currently dubbed because it was not economically viable to dub.”

Resemble AI, another company looking to capitalize on localization, takes a different approach. Its technology focuses on creating custom voices using proprietary deep-learning models that can produce realistic sounding “unique” voices. These voices are typically used in animated how-to videos, and commercials, rather than film and television shows, but the principle remains the same. These AI generated voices are more cost effective than traditional voice over or dubbing, and companies are turning towards start-ups like Resemble AI for their dubbing needs. 

ElevenLabs is taking it one step further. Its AI dubbing tool can translate speech into more than twenty languages while retaining the original speaker’s voice. The founders of ElevenLabs claim that their technology can “deliver high-quality translated audio” to producers within minutes. Impressively, the AI specifically seeks to preserve the original speaker’s voice and intonation. 

So what does the rise of AI mean for the localization industry? The most immediate impact is obvious – a lot of people will likely lose their jobs – from voice actors, to translators, to producers. The localization workflow process – where studios hire an entire local team to review, edit, translate, and dub a property for international audiences – is going to be a thing of the past – replaced with slick, cost-effective AI alternatives. 

But more broadly, generative AI represents an opportunity. Thousands of properties are never localized in certain languages or regions, and some properties are never localized at all. With AI making dubbing and subtitling so streamlined, the cost benefit now weighs in the contents’ favor, allowing it to be translated and distributed to places it never would have otherwise. Localization workflows will become more efficient and more content will become available to a broader audience faster. 

More jobs will open in the creative tech space – these tools will only become more advanced, and there is a need to provide cogent, relevant data for the machine to train on. Job opportunities in quality control and AI monitoring will expand, as more traditional roles vanish.  

While the full extent of generative AI in the entertainment industry is just beginning to be grappled with, its impact on localization is obvious and ever expanding. AI is impactful, efficient, and here to stay. The question which remains is how the industry will react to it.