Virtual production 2026: Is AI a Dangerous Industry Risk?

What’s new in Virtual Production?

Artificial intelligence has yet again found its way into another industry, and this time it’s taking the world of virtual production by storm. With the blend of physical and digital realities, the possibilities are endless.  Although there may be doubt about whether something you see online is AI-generated, rest assured that there are real people behind the prompts creating each piece.  

Trends are switching faster than ever before in the virtual production world, and the workload is starting to tower over teams who rely on the traditional way of working. Studios that have branched out to use artificial intelligence have been granted an unlimited fast-track pass to quicker production. It no longer takes several months to conceptualise an idea, as AI can now automate time-consuming tasks without sacrificing the quality of the output. 

What is Virtual Production? 

Virtual production is the incorporation of digital sets and LED screens to create a scene for live-action shots. The concept was first introduced as early as 1930 in a film produced by Fox Corporation, famously known for Fox News, etc. “Liliom” is a love story between a passionate baker, captivating customers with his charm and a young lady who fell for that charm, willing to risk it all for their romance. This movie displayed the use of rear projection, one of the first of its kind in virtual production. This is the act of screening an image behind a translucent screen for subjects to elude the viewers from thinking it’s a fake background.  

Virtual productions’ further developments include the first “Avatar” movie, published in 2009, which pushed the boundaries of conventional filmmaking by using advanced performance capture and virtual cameras. The integration of gaming engines such as “Unreal Engine” allowed for live previews due to 3D rendering tools, improving efficiency on set as adjustments can be made in real-time. Filmmakers can now obtain instant results as the CGI characters and sets display live feedback on the entire composition. This marked the beginning of modern virtual production as the post-production workflow was improved, allowing creative decisions to be made during the shoot.

The popularisation of ‘LED volume’, a curved LED screen capable of displaying photorealistic environments in real-time, in virtual production, was achieved through the series “The Mandalorian” set in the Star Wars universe. With its advancement, they were able to abolish most green screen works and achieve realistic lighting while minimising parallax. The use of virtual production greatly reduced set design costs, as being able to display complex environments onto LED surround screens prevents unnecessary budget for travelling or constructing new sets. As “The Mandalorian” is filmed on an episodic production schedule, having the ability to reuse and modify previously used environments allows for an efficient workflow and maintains visual consistency during production.

How is Artificial Intelligence changing production? 

Speaking of efficient schedules, Generative AI allows filmmakers to cut down on core functions such as creating a digital environment, which is crucial in virtual production. As users continue to contribute to machine learning, the AI will start to understand the patterns in the virtual set and produce natural-looking scenes with interactions that look realistic. Have you ever thought of scrapping an idea mid-way through production? Artificial intelligence can help you create a concept which is completely new or follows along the same plot line as your original story.  

Pre-production is one of the most challenging stages of production, as coming up with an original idea that will stick within a strict budget and meet deadline requirements is not an easy feat. Artificial intelligence can assist in developing start-up ideas that can be built upon to become an award-winning movie.

During the pre-production of the movie ‘Stevie‘, written by screenwriter Robert Marsh, based in Chiswick, a few mock-up scenes were created using AI for a visual mood board. This was done to get a feel for what actors should be cast for the movie, along with the atmosphere that’s meant to be portrayed in the film.

Real-world examples of AI in Virtual Production:

You’d be surprised at how many hit movies have used Generative AI to help with their production. A few examples are: 

• The Lion King (2019) utilised AI production to render in environments and complex animal models to make them as realistic as possible. This allowed the seamless blend between live capture and the digital world.                                                                                                            
• Top Gun: Maverick used the power of AI voice synthesising to recreate Val Kilmer’s voice after he lost his ability to speak due to cancer treatment. Sonantic was able to recreate Kilmer’s voice by using hours of recordings, which produced an accurate replica of his voice for the character Iceman.                                                                                                                                                  
• The Brutalist sparked debate in using AI tools, as the concept of the movie was from the Brutalist architectural movement emerging in 1950s England, back when technology of that kind was not available. A native Hungarian-speaking editor, Dávid Jancsó, fed his own voice into the voice refinement technology to subtly correct the Hungarian pronunciation of the lead actors’, Adrien Brody and Felicity Jones, dialogue to make it sound more authentic while still preserving their original voices. To gain inspiration for the architecture during that brutalist era, production used generative AI to create quick mock-ups, which were used for reference, to later be drawn up by hand to align with the film’s aesthetic and period-specific style.                                                                 
• Dune: Part Two took advantage of Generative AI and created futuristic cities and vast deserts for its background landscapes. This was mainly done in virtual production using Unreal Engine to enhance features for efficiency on schedule and quality, as the detailing would’ve taken too much time to do traditionally.
• Late Night with the Devil experimented with Generative AI in virtual production and decided to use it to make title cards for the talk show in the movie. There were three cards made in total, which were intended to enhance the 1970s feel of the movie. However, similar to the brutalist, this did spark backlash. The production was for an indie film that relied more on practical 70s-style effects, but with the use of AI, it didn’t feel authentic after the media discovered the truth.