Tech Policy at Duke University

The 2025 Academy Awards saw the first glimpses of how AI technology is changing the filmmaking process. It also showed just how audiences are responding to AI.

The Brutalist was nominated for 10 Academy Awards that year, eventually winning three, including Best Actor. But the movie’s legacy won’t just carry three Oscars, it will also carry a controversy about its use of AI. During the film’s awards campaign, editor Dávid Jancsó spoke with RedShark News, a video technology outlet. In the interview, Jancsó revealed that the production used AI voice-generating technology from a company called Respeecher to enhance the voices of the lead actors. Both actors, Adrian Brody and Felicity Jones, are English speakers who were tasked with speaking Hungarian in some scenes. Jancsó’s voice, speaking Hungarian, was fed into an AI model, which then modified Brody and Jones’ Hungarian accents.

Public opinion on AI in film

The Brutalist wasn’t the only Oscar-nominated film to use Respeecher. Emilia Perez, which was nominated for 13 Academy Awards and won two, used the technology to extend the vocal range of one of its lead actresses. The use of AI in these productions sparked online debate and backlash, with debates on whether AI enhancements undermine the actor’s performance.

Despite the controversy, public opinion on the use of AI in film isn’t clear-cut. An iHeartMedia study found that 75% of respondents don’t want AI in their media. In a YouGov UK poll, 48% of respondents said that it is acceptable for filmmakers to use AI in the filmmaking process; 39% of respondents said it was unacceptable. Specific use cases were considered acceptable by a majority of respondents, such as translating and dubbing the film into different languages and generating visual effects (VFX). However, a plurality of respondents agreed that films using AI should not be eligible for major awards.

A year after The Brutalist’s controversy, scrutiny of the use of AI in film continues to be discussed on social media. Recently, the upcoming movie Stop! That! Train! faced backlash for the perceived use of AI, including calls to boycott the film. Speculation of the use of AI came from a Letterboxd review of the movie by an early reviewer, who claims that shots in the movie appear to be AI and that Acme AI—a VFX AI company—is credited for the movie’s visual effects. Director Adam Shankman released a carefully worded statement that there are “ZERO shots conceived by AI” and that every shot “was made by human hands.”

The evolution of film technology

The controversy over Stop! That! Train! highlights how production technology has evolved over time with the rise of AI technology. Film technology has progressed immensely, with digital tools significantly impacting production costs and processes through efficiency gains. While early advances in film technology brought movies with sound and color, the shift from analog to digital production introduced new visual effects and editing capabilities. Computer-Generated Imagery, or CGI, gained popularity starting in the 1970s, allowing more complex and sophisticated images to be incorporated into live-action movies. The technology can create 3D models, including human faces, backgrounds, landscapes, and animals.

The capabilities of CGI sound the same as AI, and it is becoming increasingly hard to disentangle the two technologies. AI and machine learning are used to enhance, automate, and integrate CGI more efficiently. AI can more effectively process motion capture, reduce distortions in the data, enhance visual simulations, and assist animation. Rotoscoping, a post-production process that traces elements of live-action footage to use as a base for CGI, can be done far more efficiently using AI.

Some use cases of AI in VFX predate the audience’s awareness of the technology. The acclaimed film Everything Everywhere All at Once used technology from Runway AI to automate rotoscoping; the film was released in March 2022, preceding the release of generative AI tools like ChatGPT, which launched in November 2022. Everything Everywhere All at Once largely escaped scrutiny for its use of AI, perhaps due to the timing of its release and clear use of practical effects. The movie may also have benefited from the public’s lack of awareness of generative AI capabilities. According to the YouGov poll, audiences are far less willing to accept AI in filmmaking when it is used to generate new characters and footage, or otherwise replace actors and human creatives.

The case for AI production technology

AI’s efficiency benefits are clear, especially for a cost-intensive industry. Janscó and the Everything Everywhere All at Once filmmakers acknowledged the cost and time-saving benefits of the use of AI. More recently, Martin Scorsese, one of the most acclaimed directors of all time, backed an AI image generation start-up he used for storyboarding his new movie. Use cases that speed up the overall production process allow for shorter production cycles and reduce the costs of production, which have benefits for filmmakers and consumers alike.

How Hollywood should respond to AI in filmmaking

The public remains wary of the use of AI in filmmaking, especially when it is used to generate elements that would otherwise be created by humans. Despite this, film festivals and major studios are beginning to roll out content entirely generated by AI, such as an upcoming film premiering at the Tribeca Film Festival and several children’s programs from Amazon’s MGM Studios.

In anticipation of the increased use of AI in filmmaking, the Motion Picture Association should introduce an AI rating system as a supplement to its existing film rating system. The current film rating system categorizes films based on their suitability for different age groups. An AI rating system would create similar categories or tiers based on the level of AI use involved in a production. A film that uses AI only for automating repetitive production tasks might receive a lower AI use rating than a film that generates entire images or sequences using AI. This would allow audiences transparency about the level of AI used in a film, enabling consumers to choose to engage with content they feel comfortable with.

Awards bodies like The Academy of Motion Picture Arts and Sciences and the Golden Globe Foundation should also consider the level of AI use in their awards nominations process. The Academy has already clarified that only roles demonstrably performed by humans will be considered eligible for acting awards and that they will scrutinize the nature of generative AI use and human authorship in films. With an AI use rating system, award bodies can be more explicit about the level of acceptable AI use for nominated films. These awards are meant to celebrate creativity and execution. Establishing parameters for the acceptable use of AI ensures that there is an appropriate level of human authorship and a more equitable playing field for Hollywood’s most acclaimed work.

This is the final blog in a three-part series exploring AI’s impacts across the film lifecycle. The series explored a range of policy issues related to the film industry, including taxation, government incentives, labor market impacts, and public perception.