AI Is Now Censoring Movies Without Filmmakers’ Consent
The world just witnessed a new low in film censorship. The Australian horror film Together, starring Dave Franco and Alison Brie, was digitally altered for its release in China. A gay couple at the heart of the story was rewritten into a heterosexual pairing using AI face-replacement technology. One actor’s face was literally swapped out to appear as a woman, and references to a same-sex relationship were stripped away. What makes this worse is that the edits were done without the knowledge or consent of the filmmakers or the global distributor, Neon. It was not just censorship. It was cultural manipulation by machine.
This case has sent shockwaves through the creative industry because it highlights a dangerous precedent. Traditionally, censorship has meant cutting scenes or muting dialogue. Filmmakers could at least recognize what was removed, even if they disagreed. But AI allows for a seamless, undetectable kind of distortion. In Together, a wedding scene was left intact — except that one groom was no longer a man. The story was made to look like it had never been about two men in love. To audiences who had no idea what was changed, the deception was complete. That is not editing. It is erasure.
Neon immediately condemned the unauthorized edits and demanded that distribution in China cease. The film was pulled from theaters, a bold move in a country where foreign releases are carefully vetted. Surprisingly, it was not only Western critics who took issue. Chinese audiences themselves, upon noticing discrepancies between the official trailer and the version in theaters, criticized the changes on film platforms and social media. Many described it as distortion and misrepresentation, proof that viewers do not want their stories reshaped into lies.
The problem is larger than one film. AI has become the new censorship tool. Where governments and distributors once relied on blunt scissors, they can now deploy algorithms that rewrite identities, sanitize sexuality, and even change the intent of art. Last year, Chinese theaters digitally added a black dress to Florence Pugh’s nude scene in Oppenheimer. That edit was crude but obvious. Now, AI makes the changes invisible. A gay man’s face becomes a woman’s, a kiss vanishes, and an audience is left thinking that the story was always meant to be straight.
This raises urgent questions of consent. International copyright law protects not only the financial rights of filmmakers but their moral rights as well. Directors, writers, and actors are entitled to object when their work is distorted or mutilated. Replacing an actor’s face with a digitally generated one violates not just the story but the actor’s very identity. It turns performance into raw data that can be manipulated without permission. The line between artistry and deepfake has been obliterated.
The implications go far beyond film. If AI censorship is normalized, books could be rewritten to remove politically sensitive words, music videos could be altered to erase provocative imagery, news broadcasts could be “corrected” to align with official narratives. What begins as a wedding scene in a horror film could escalate into the rewriting of history, delivered through seamless visuals that audiences may never question.
The betrayal here is not only against filmmakers but against audiences. People buy tickets believing they are seeing the director’s vision, the actors’ performances, the story as it was meant to be told. Instead, they are handed a counterfeit — cinema that has been engineered to meet state or distributor standards. That is not storytelling, it is propaganda by machine. And when it happens without disclosure, audiences are robbed of their agency to decide for themselves.
There is no denying that many regions enforce strict cultural sensitivities. The Middle East, parts of Asia, and China have long histories of restricting content. But the correct path is negotiation. Studios may choose to agree to cuts, or they may refuse release altogether. What is unacceptable is unilateral, AI-driven modification without consent. That is not cultural adaptation, it is creative theft.
The global industry must treat this as a wake-up call. International standards must be strengthened, and unauthorized AI edits should be met with penalties and distribution bans. Technology that can be used to empower creativity cannot be allowed to become a silent eraser of inconvenient truths. If unchecked, AI will not just be a creative tool. It will become the ultimate censor.
