What happens to an industry's workforce when the camera no longer needs to capture the truth of a face, a voice, or a body at all?

For years, the illusions created for Indian films were done by hand by craftspeople. With the help of make-up artists who added age to actors using latex and shading, junior artists filling in background crowds, dubbing professionals who spoke for characters in dozens of languages, and colourists/VFX technicians working tirelessly on a single frame.

Now that the craft economy is slowly being absorbed into prompts and pipelines, there's no need to rely on all those craftsmen anymore, because artificial intelligence has outgrown its original purpose as an ancillary function of pre-visualisation. This article explores what that shift means for the junior artists, dubbing professionals, and VFX technicians being absorbed into prompts, and whether India's unions are ready for it.

From Rotoscoping To Retouching, In Real Time
Sidharth Meer is one of many colourists who have worked on projects such as Homebound, one of India's entries for the 2025 Oscars. For example, when working on Jigra, Sidharth's team used face-tracking tools to isolate and enhance Alia Bhatt's eyes rather than spending hours manually rotoscoping the footage frame by frame.

Sidharth has spoken openly about how this changes artists' labour and workflow: tasks that would previously have taken countless hours to track facial elements manually can now be completed in a fraction of that time. It's not just an "extra" creative flourish; it's a direct replacement for what a group of rotoscopers previously completed over several days.

The same logic runs through Dibakar Banerjee's Love Sex Aur Dhoka 2, widely regarded as India's first significant commercial deployment of AI in a climactic sequence, and Studio Blo's IRaH, which reportedly used AI for roughly 60 per cent of its visuals, de-ageing actor Rohit Bose Roy from 52 to 22.

These types of experiments happen regularly. In the Hollywood Reporter's AI Issue, Mumbai AI Studio believes that most Indian cinema (over 80%) has incorporated AI into its pre-viz processes. At the same time, full-length features can be produced via AI-powered pipelines that reduce what would usually take years to a few months.

A Decade In The Making
Bollywood has long utilised de-ageing, with examples such as Shah Rukh Khan's use of it in both Fan and Jawan (respectively) and Salman Khan's multiple decades through digital means in Bharat. The technological tools that have enabled this are vastly different from those in use over the last decade in terms of budget and accessibility.

De-ageing sequences required significant amounts of money, time, and (often) special-effects studios until recently, so it was an extremely ambitious approach. Compared to the technology used in the past, the tools now available to filmmakers can accomplish tasks much faster and at a lower cost than ever before.

The popularity of de-ageing has since spread to multiple forms of Indian cinema, including but not limited to Tamil superstar Rajinikanth's de-ageing sequences in Coolie, as well as Malayalam superstar Mammootty's de-ageing in Rekhachithram, which was done using more than 1000 archival photos, indicating that de-ageing is no longer a thing limited to Bollywood but one of more interest and acceptance throughout India. Rekhachithram has since become one of the biggest successes in Malayalam cinema, indicating that the audience is not only accepting but also rewarding its filmmakers at the box office for these techniques.

Not every star is convinced the trend is as widespread as headlines suggest, however. Actor R. Madhavan has publicly pushed back against rumours that he has used de-ageing technology in his own work, noting that such tools remain costly and are not yet a default option, even for established names. His comments are a useful corrective: AI de-ageing is spreading fast, but it has not yet become the industry default it is sometimes assumed to be, and budget remains a real constraint outside the biggest tentpoles.

The Dubbing Booth Is The Canary In The Coal Mine
The most apparent facet of this transition is the de-ageing of individuals. Still, it is in dubbing that the worst economic impacts of this aforementioned shift will undoubtedly be felt. India's dubbing industry is an extremely large, largely freelance-based ecosystem that serves a film industry spanning over 10 languages.

The business rationale being used against the dubbing industry is very straightforward. If AI-based dubbing results in audio files that are nearly indistinguishable from those recorded by human voice artists, and can provide a single film dubbed into 5-6 different languages at a fraction of the price that is currently being charged for traditional dubbing, then there will no longer be a business case for having many human voice actors.

Currently, this is no longer theoretical or hypothetical. There has already been significant movement towards the use of AI voice technology by the vast majority of commercial TV and video advertising brands in India, and this trend is beginning to extend into narrative programming.

For instance, for War 2, Yash Raj Films used lip and gesture-modification technology to make the originally filmed Hindi dialogue appear as if it were originally spoken in other languages, restoring a co-star's original voice and distributing the facial video for the regional-language version – even though that co-star had never delivered that line in that particular language while performing.

Major streaming platforms are simultaneously rolling out their own voice-cloning and lip-sync systems to localise entire content libraries at speed. One veteran voice artist who has dubbed extensively for streaming platforms has put the warning starkly: if AI fully takes over, the dubbing profession as it currently exists is finished.

Junior Artists, Technicians, And The Missing Safety Net
People most affected by this shift aren't the actors whose faces are being re-aged. They're the artists who have done rotoscoping, the junior voice actors, the extras who were used to fill in the gaps between what the camera recorded and what the final movie needed; the people who did the first set of VFX and the people who were responsible for integrating all those elements to create a finished product. As artificial intelligence continues to take over the gap-filling part of this area, there is a question of how quickly this will happen and whether anyone will be representing these types of jobs during the transition period.

In contrast to Hollywood, the WGA and SAG-AFTRA strikes of 2023 are largely about securing protections through contractual guardrails governing the use of AI — specifically, that performers must consent to the digital alteration/duplication of their likeness or voice.

India does not have any comparable industrial action laws or union structures to effectively lobby for such protections for workers. The discussion surrounding India's AI film industry boom has consistently highlighted that the country lacks empowered industrial unions to lobby for caution and that there is a lack of national regulatory legislation on AI and employee protection.

Generally speaking, most contracts in the Indian entertainment business grant studios the right to use finished products in any format or technology available today and in the future; however, very few contain language specifically addressing how to change a performer's or actor's likeness once the performance has been recorded.

This became apparent when Eros International released an 'AI-altered' version of the film Raanjhanaa with a newly created ending. Despite a concerted public outcry from the film's director and lead actor (both of whom argued that this change violated the integrity of the film), ultimately, legal analysis indicated that Eros International probably had the right to alter the film, given that in most 'work for hire' agreements, the studio is the first owner of all created material.

Since then, there have been concerted efforts to draft directors' contracts to include clauses requiring the director's consent before any future alteration to a finished product, with very little implementation of such clauses in practice.

Are India's Unions Ready?
Not yet, to answer directly. India's entertainment industry contains unions such as the Federation of Western Indian Cinema Employees (FWICE), which include a wide variety of technicians and junior performers. Still, historically, they have focused on wage negotiations, working conditions, and credits rather than directly addressing the fast-moving threat posed by generative AI.

Currently, there are no Indian production agreements that provide AI consent and compensation frameworks similar to SAG-AFTRA's, nor any legal frameworks to enforce them. It will take time for these protections to become established through the courts, and there are still ongoing copyright cases in which courts are deciding whether or not AI was trained on copyrighted material.

What this means is that those building India's AI film production pipelines are not predominantly bedroom auteurs democratizing access to filmmaking, despite all the references made to that narrative.

Large conglomerates and well-established production companies and streaming services are investing heavily in their AI capabilities and have both the funds to develop proprietary technologies and the capacity to integrate smaller, more specialised entities into their larger systems. This concentration of capabilities, combined with the lack of concentrated bargaining power among displaced members of the workforce, creates the very kind of imbalance in the labour market that organised labour exists to rectify elsewhere in the world.

The Takeaway
None of this means Indian cinema's crew economy disappears overnight. Directors and cinematographers describe AI mainly as a tool for ideation and pre-visualisation, solving problems that traditional techniques struggle with, such as visualising an abstract idea before a shot is taken. Many insist judgment remains irreplaceable, and that AI amplifies production rather than replacing the people directing it.

The perspective of the person performing rotoscoping/voice dubbing is different from theirs; the issue here is not whether or not the tech works, but whether the pace and lack of union representation in this industry create an environment where individuals do not lose credit for their work. Jigra, Rekhachithram, and War 2 have shown that the tech works.

Either way, India is no longer watching from a distance. It is the world's most consequential live experiment in AI replacing film labour.