The one-second shot that exposed Hollywood’s AI secret
Ben Affleck’s Netflix thriller shows why hiding where artificial intelligence was used costs more trust than the shortcut ever saved.
Sign up for my Substack daily AI newsletter here.
Follow me on LinkedIn: https://www.linkedin.com/in/johanosteyn/
A stunt shot in Animals, the Netflix thriller that Ben Affleck directed and stars in, lasts less than a second. A kidnapped boy is pushed down the wall of a Los Angeles canal and lands in Affleck’s arms. Viewers have replayed it in slow motion, and South Africa’s Daily Voice judged the AI-generated movement “strikingly swak”, seeing the scene as a cost-cutting choice over skilled craftsmanship. The shot matters less than what surrounds it. Affleck founded the AI company whose tools were tied to the film, and he has declined to say where AI was used. This article looks at why secrecy, rather than the technology, became the story, and what that means for any organisation putting AI in front of its customers.
CONTEXT AND BACKGROUND
This year, Netflix bought InterPositive, the AI post-production company Affleck founded, so Animals was always going to be watched as a showcase. Before release, Affleck confirmed the film used AI, called it “a very human movie” and declined to give details, comparing disclosure to a magician telling the audience, “I want you to know there’s another bunny under here.”
The commercial picture complicates the backlash, although many viewers probably pressed play before the clip spread. By 17 October, eight days after release, Animals was the most-watched film on Netflix in 78 countries, according to FlixPatrol figures. Netflix also says generative AI was used on roughly 300 titles in the first half of 2026, and co-chief executive Ted Sarandos says 17 minutes of footage in one documentary series were produced about twice as fast and at roughly half the cost.
INSIGHT AND ANALYSIS
I have previously written about this when Netflix bought InterPositive, arguing that AI is rewriting Hollywood’s cost curve and will democratise who gets to make films. I still believe the cost curve is shifting. Animals shows the condition I underplayed: lower costs widen access only when the result is good enough, and audiences now judge AI work by its weakest frame.
Danny Boyle offers the instructive contrast. His film Ink discloses its use of generative AI in the credits, and Boyle has said there are “about 30 seconds” of it, used to animate old photographs and to create mice. He still drew criticism, but he set the terms of the conversation. Affleck’s silence handed those terms to the internet.
Audiences increasingly expect to be told. In a British Council survey of 2,001 people aged 16 to 24 in the United Kingdom, conducted by Censuswide in September, 90% said productions should be required to disclose AI use, and 46% would be less likely to watch if they knew AI had been used extensively. It covers one country and one age group, but the direction is clear.
The strongest counter-argument deserves weight. The disputed shot is a fraction of a second, conventional visual effects fail too, and the viewing figures suggest no commercial harm. Disclosure rules that apply only to AI could also single out one tool among many. That is fair as far as it goes. Audiences have long forgiven clumsy wire removal or rubbery digital doubles, however, because those flaws carried no wider fear. Generative AI arrives with anxieties about jobs and authenticity attached, so its mistakes are read as a statement about how the work was made. Trust is rarely lost first in the viewing figures. It is lost in the story people tell, and here that story became a founder showcasing his own technology without saying where.
IMPLICATIONS
In South Africa, the film and television industry has spent much of 2026 fighting delays in the government’s production incentive, and Parliament’s trade committee welcomed a joint plan in August to clear the backlog, with a progress review due in November. Producers under such financial pressure will find cheap generative tools tempting. Used well, they could help small local teams compete with far larger studios; used carelessly, they produce the work audiences mock.
Any organisation putting AI-generated adverts, voices or images before customers faces the same test. Boards should ask three questions. Who signs off AI output before it is published, and against what quality standard? Is disclosure policy decided in advance, or improvised after a backlash? Where the people promoting a tool also have a stake in it, who checks the work independently?
CLOSING TAKEAWAY
The lesson of Animals is not that AI has no place in creative work. It is that concealment turns a small technical flaw into a question of trust. A shot lasting less than a second became a global talking point because audiences felt they had been asked not to look. Organisations will keep using AI, because the savings are real. The ones that keep their customers will set a clear quality floor, say plainly where machines were involved, and keep skilled people responsible for the final result. Audiences forgive imperfection far more readily than they forgive being kept in the dark.
Author Bio
Johan Steyn is a prominent AI thought leader, speaker, and author with a deep understanding of artificial intelligence’s impact on business and society. He is passionate about ethical AI development and its role in shaping a better future. Find out more about Johan’s work at https://www.aiforbusiness.net
Comments