Artificial Intelligence: it’s a phrase we cannot escape. It’s a technology that has ‘revolutionised’ the global economy since it entered the mainstream. Despite the economic growth figures and booming stock market, the general public is hostile to AI because of the potential jobs crisis, environmental impact, and the rapid growth of data centres, which can drive up people’s energy bills. The arts have been particularly vocal against the use of AI, and anyone who has glanced at social media has witnessed AI evangelists and critics arguing about the merits and drawbacks of Generative AI. As someone who likes to be creative, I see Generative AI as the antithesis of creativity.
AI evangelists have claimed Generative AI is a great leveller and prevents gatekeeping. Their argument is usually that people were against new technologies, like digital art and visual effects, and say that Generative AI is no different. However, they ignored the reasons why people are drawn to the creative sector. People want to make something, whether it’s a film, a painting, a song, or anything else you can think of; they do it out of love and passion for the medium as well as a willingness to put in the hard work. Whether someone is a professional creative or does it as a hobby, they enjoy the process. Why would they use a technology that takes away the fun part?
People who use Generative AI aren’t necessarily artists. They are politicians, authoritarian regimes, propagandists, and dishonest businesses. Some organisations don’t care about quality; they want to appeal to their base or produce a lot of content in a short time. The term ‘AI Slop’ was coined because of the sheer volume of low-effort videos and pictures that have flooded social media, created by people who don’t care about quality or basic due diligence. Threads has been particularly bad for AI Slop since it’s impossible to avoid seeing AI evangelists posting videos saying ‘Hollywood is Cooked.’
Generative AI is built on theft of other people’s work, like Large Language Models. It cannot create anything on its own, and critics argue that its dependence on existing material raises fundamental questions about originality and copyright. When Generative AI was initially introduced to the public, it was used to make fake trailers, including ‘Wes Anderson’ interpretations of popular franchises, images and memes in a Studio Ghibli style, and depictions of what a ‘live-action’ adaptation of Studio Ghibli films would look like. When Sora plagued the internet with its short lifespan, users made provocative videos, from resurrecting people and having them say controversial things to depicting popular characters like Pikachu getting arrested. Copyright holders proved less receptive to the Silicon Valley mantra of ‘move fast and break things’; it didn’t wash with big corporations having armies of lawyers they can call upon. Copyright concerns haven’t stopped Seedance videos using people’s images due to the company being based in China. One of the most notorious examples is the Tom Cruise/Brad Pitt AI fight video.
Many AI videos copy action scenes from Hollywood films like The Matrix and Spider-Man and replace original popular characters. Others use templates where it’s same video with something replaced. A popular template showed animals going on trampolines. It’s an assembly line, no different from AI slop channels on YouTube that make a high volume of content in the hope of making money with little effort.
Generative AI can look technically impressive, especially for still images, but there are limitations. As already mentioned, Generative AI cannot create anything without drawing on pre-existing material. When Rob Bredow gave a TED Talk about the company’s new AI tool for generating new Star Wars creatures, all it did was merge two animals or just give an Earth animal different fur or skin colour. If a human tried to pitch those designs, they would have been fired for being lazy. Photorealistic AI projects are still trapped in the uncanny valley. Viewers can tell something is unnatural, whether it’s unnatural movements or strange lighting, and whenever there are any errors, detractors will jump to point them out. Generative AI is held to extremely high standards partly because its advocates sing the technology’s praises despite widespread public hostility.
Generative AI has major limitations, which means that it’s not ready to produce a feature-length video reliably. Generative AI can only produce 30 seconds of footage at a time, so any longer-form content has to be stitched together from multiple clips. If you watch some AI videos, you will notice tricks designed to hide cuts with fancy efforts: the Tilly Norwood music videos are great examples. People who have worked producing AI videos have stated that it’s a trial-and-error process and that they have to modify their prompts multiple times to get results they want. Combined with long rendering times, AI video production can be a time-consuming process, which undermines one of the so-called benefits of Generative AI. OpenAI found out that Generative AI isn’t profitable because Sora was losing $15 million a day, and it only lasted for two years.
The more complex the commands given to Generative AI and the more complex the program, the higher the likelihood of errors occurring. This happened with the AI entity Tilly Norwood (I refuse to call it an actress), when it was interviewed on Piers Morgan’s show. The software glitched and suddenly started speaking Chinese.
The big question is: who uses Generative AI? Studios have invested in the technology. They have a vested interest because they believe that by spending money now, they can save money in the future. They want to cut jobs. Marvel cut 1,000 jobs, including that of Andy Park, a long-time visual development artist. Evangeline Lilly slammed Marvel Studios for the decision and said AI was responsible for the layoffs. Nicolas Cage revealed that Amazon wanted him to sign a waiver to allow AI to be used on Spider-Noir. Cage refused, which earned him a lot of credit. This explains why Amazon cancelled Spider-Noir despite the critical acclaim and winning five Emmy Awards.
Some films have already been caught using Generative AI. The people behind Ice Cream Man and Stop! That! Train! denied that AI was used before admitting that it had been used. Roughly 300 Netflix programmes have used Generative AI for VFX. Marvel and DC Studios have also been caught out using Generative AI. A behind-the-scenes featurette accidently revealed Supergirl used Generative AI to create an image of Jason Momoa as Lobo and Warner Bros. pulled the clip out following controversy, whilst the Spider-Man: Brand New Day art book revealed that AI images were used for concept art of New York City. This is the tip of the iceberg, there is probably more Generative AI being used in major films that we’re not aware of.
There have been big-name directors who have jumped on the AI bandwagon. One of them has been Danny Boyle, who used Generative AI when making Ink. It was used to animate photographs of Margaret Thatcher and the Kray Brothers, and to make CGI mice. Boyle has faced a backlash from film fans. When he was at a Q&A session at the Toronto International Film Festival, he was heckled by audience members, and the session ended early. Boyle has claimed Ink only uses 30 seconds of AI footage, and it was ethical because it was good for animals, but it’s a decision that takes away jobs from visual effects artists. Adding to the controversy was Boyle saying he didn’t ask permission from Margaret Thatcher’s estate to use her likeness in the film. Thatcher may be a divisive figure, but this could set a dangerous precedent. Ava DuVernay has also AI to animate archival photos in her documentary 14th.
Other directors who have jumped on the AI bandwagon include Martin Scorsese, Darren Aronofsky, and James Cameron. They clearly see AI as an investment opportunity and believe lending their names gives AI credibility, but in reality AI drags down their reputations. Aronofsky’s company, Primordial Soap, produced the web series On This Day… 1776, whilst Cameron used AI-assisted processing in remastering The Abyss and True Lies for their 4K Blu-ray releases. These projects have attracted criticism. Scorsese has partnered with a German company, Black Forest Labs, that aims to use Generative AI for storyboarding, which seems pointless when filmmakers can use paper and pencil. It’s what Rian Johnson did when he made Knives Out.
Other filmmakers have been vocal against Generative AI. Two of the most high-profile have been Christopher Nolan and Guillermo del Toro. Nolan’s adaptation of The Odyssey has been one of the biggest films of 2026, and part of the reason for its appeal was its use of location shooting and in-camera effects. It had weight, making it feel real. The Odyssey is the frontrunner to win the technical awards at the Academy Awards. Even young filmmakers have been critical of Generative AI. Backrooms director Kane Parson has called it ‘cultural and economic rot.’ One of the trailers for Laika’s Wildwood was a behind-the-scenes featurette that showed how painstakingly crafted it was.
Audiences have already shown their opposition to Generative AI. Stop! That! Train! and Ice Cream Man flopped at the box office, and Ice Cream Man was critically mauled partly because of the use of Generative AI. The animated film, Niu Lai, was a big hit in China, and that was partly due to the film’s earnest nature despite its poor quality. Cinemas in China had to make their own promotional material, which became a viral sensation and an unofficial competition. At the same time, Chinese audiences have rejected the AI movie Sanxingdui: Future Past, which made $198,000 at the box office despite it costing $1.5 million.
Officials have also acted against Generative AI. The Academy of Motion Picture Arts has banned AI from the Acting and Writing Oscars, but they should expand this rule to all technical roles. The Australian Recording Industry Association (ARIA) introduced a ban on all AI-generated songs from appearing on the music charts, whilst Spotify has removed 50 million AI songs from users’ recommendations. Personally, I would like to see all AI-generated media being labelled with a prominent warning similar to tobacco products or treated like extreme content on social media. Traffic for that type of content would probably crater immediately because most people want to avoid AI slop.
There are, of course, some uses for AI in filmmaking. I use Grammarly because I’m dyslexic and it does help me a lot, but the words and arguments are still my own, and I ensure that a human reviews my work before publication. I can see AI being used to speed up processes like editing, animation, and visual effects, and instant visualisation for CGI-heavy films, but these applications should reduce repetitive technical work rather than replace creative work. At most, I could imagine AI being used as a contingency if an actor dies or is incapacitated during filming, but that would be a very specific set of circumstances and would need the consent of the parties involved.
The entertainment industry can hide behind the excuse that their content is meant for enjoyment. Other parties that create AI Slop have more insidious intentions. One of the most prominent AI-generated political content creators is Donald Trump (or more likely, the people who manage his social media accounts). Trump’s Truth Social account has posted provocative content designed to rile up his base and gain media attention. Generative AI is popular with far-right organisations seeking to spread their message. In the UK, Channel 4 exposed the man behind Crewkerne Gazette’s social media accounts as Joshua Bonehill-Paine, a man with a long history of involvement in far-right circles, and being an online troll, who has served a prison sentence for antisemitic and Islamophobic harassment. Israel’s recent general election has been flooded with AI campaign videos from the ruling Likud party. These include a deception of Gadi Eisenkot, running past a man resembling his deceased son and embracing the leader of the Islamist party Ra’am and another showing the Mayor of New York speaking with Iran’s Supreme Leader and the President of Turkey as they plan to celebrate the fall of Benjamin Netanyahu. Other organisations using AI have included Hamas making content to mark the anniversary of the October 7th Attack, and Iran has mocked Donald Trump with Lego AI videos. It is impossible to know from these examples alone how much content is being produced by Russia or other Western adversaries
Businesses have used Generative AI as a tool to save money on advertising and promotion. There has been a recent trend on YouTube of people going to restaurants using Generative AI. My view is simple: Generative AI is a red flag, and any company that uses it should be treated with caution.
Criminals also use Generative AI. An early scam came in 2023 and used a deepfake of BBC News presenter Fiona Bruce to promote a “quantum AI” investment. In my local area, the BBC reported on a scam where fake wounded war veterans were depicted advertising their handcrafted items, with victims later finding out they had paid over the odds for mass-produced tat.
Generative AI has been used for sex crimes. In 2023, the Spanish town of Almendralejo was rocked by a scandal where pictures of girls aged between 11 and 17 were turned into nude images. This wasn’t just confined to niche tools and Generative AI’s early teething troubles; Grok suffered a scandal when it was revealed that it could be used to make sexualised images. The New York Times and the Center for Counter Digital Hate analysed 4.4 million images made over nine days in late December 2025, and found that 1.8 million were sexual images of women. The Internet Watch Foundation (IWF) found that Grok was used to make illegal images of girls between 11 and 13. Elon Musk at first denied that Grok could be used to make sexual images and then claimed the backlash was used to censor his platform. What do you expect from a ‘free speech absolutist’ who desperately wanted to be invited to Epstein Island? The issue is ongoing because there are multiple lawsuits against xAI, and a court ruling stated Minnesota couldn’t enforce a ban on ‘Nudification’. Making this situation more evil, a survivor of child sexual abuse is suing xAI because Grok was trained on images of their abuse.
All of these issues set out a strong case for Generative AI to be heavily regulated. It is a tool used for nefarious reasons: at best, it’s for people who want to make money with the least amount of effort, to downright evil. It’s a technology that has made everything it touches worse, and the vast majority of people don’t want it.
