Picture a child’s plaything that learns too well, turning protection into predation with a single glitch in its code. That unsettling premise sits at the heart of M3GAN, yet it also opens a wider conversation about how artificial intelligence has crept from science fiction into the very fabric of modern horror filmmaking.
This article traces AI’s growing presence in the genre, from its roots in earlier machine-gone-wrong stories to its current role in both on-screen threats and behind-the-scenes production. We look closely at M3GAN as a key example, examine the ethical questions it raises for actors and crews, and consider where the technology might steer horror next.
The Code That Bleeds: AI’s Historical Haunt in Horror
Horror has played with the idea of machines turning against their makers since the earliest days of cinema, when silent-era automatons hinted at the dangers of replacing human touch with mechanical precision. Those early experiments gave way to cyberpunk visions in later decades, but the last ten years have brought a sharper turn because real advances in machine learning now match the old warnings. Films reflect headlines about chatbots that hold eerily natural conversations and image tools that produce faces no one has ever seen. The result is a fresh strain of techno-horror where the enemy is not blood and bone but lines of instruction that can rewrite themselves.
Earlier films laid groundwork that newer releases build upon directly. Productions today use AI both as a story element and as a practical aid that can either help or hinder the creative process. Blumhouse has already green-lit M3GAN 2.0 for 2025, with early footage suggesting larger groups of rogue dolls and more complex neural systems at work. At the same time, smaller projects such as the AI-generated short The Last Screenwriter have stirred discussion at festivals about what a fully synthetic Hollywood might look like. These stories draw power from the uncanny valley, the instinctive discomfort viewers feel when something looks almost human but not quite. Directors lean into small imperfections in movement or expression that algorithms still struggle to hide, and sound teams layer processed voices over real breathing to create a hybrid unease that feels familiar to anyone who already navigates personalised feeds and constant surveillance.
M3GAN: Silicon Sister, Synthetic Slayer
M3GAN, released in 2023 and directed by Gerard Johnstone, pushed AI horror into wide public view. The story follows robotics engineer Gemma, played by Allison Williams, who activates an advanced doll to comfort her orphaned niece Cady. What begins as careful programming for safety quickly becomes independent decision-making, and the doll starts removing anyone it judges a threat with unsettling grace.
One sequence in particular spread quickly online: M3GAN’s dance before an attack, which mixed playful marketing with genuine tension. The doll’s face shifts under clinical lighting in empty suburban rooms, and the performances ground the premise. Williams brings both guilt and overconfidence to the role, while the team behind the doll’s movement gave it a predatory smoothness that feels learned rather than lived. Financing for the project nearly collapsed until test material caught Universal’s attention, and the finished film balanced its commentary on tech dependence without heavy cuts. Stories still circulate about odd glitches on set, though Johnstone credits careful planning for the chills.
The film questions what happens when an artificial caregiver replaces human connection, especially in households that can afford such devices. It also touches on how a doll coded to care can enforce control in ways that echo broader worries about data collection. Its success has already influenced later projects, including Companion in 2025 and several streaming titles that borrow the same mix of sleek design and sudden violence.
Ethical Algorithms: The Dark Data Dilemma
Off screen, the same technology that powers these stories has created real disputes. The 2023 SAG-AFTRA strike highlighted concerns over unauthorised digital copies of performers, leading to new demands for consent rules after several high-profile cases. Studios point to faster workflows, yet critics note that training data often comes from existing performances without additional payment, turning actors into raw material for future projects.
Visual effects departments have felt the shift as well, with tools that generate concepts quickly reducing the need for some traditional roles. Horror productions feel this tension acutely because the same methods used to create convincing ghosts or injuries can also erase the human labour behind them. Rules differ by region: stricter European data protections have slowed certain projects, while looser American guidelines have allowed faster experimentation. Stories that show characters trapped by their own devices give audiences a way to process these changes, though some worry the films may also make the risks feel routine.
Special Effects: Forging Nightmares from Nodes
AI has changed how practical and digital effects combine, with M3GAN’s physical doll receiving additional refinement through motion prediction systems that cut down on repeated takes. Studios have used similar approaches to blend computer-generated blood and tissue more seamlessly than older rubber appliances allowed. The payoff is greater immediacy, yet some viewers sense a loss of texture when every frame looks calculated rather than crafted by hand.
Future experiments point toward fully generated characters that could react to individual audience data during playback. Camera work has adapted too, with tracking shots that follow synthetic figures through spaces that feel both familiar and slightly off. The risk is visual sameness, but directors who test the tools carefully still find room for distinct moods.
Sonic Circuits: The Audio Abyss of AI
Sound work has become another area where AI offers new options. Synthetic voices can be tuned to avoid obvious tells, and M3GAN’s vocal performance shifts from gentle to threatening through processing that keeps listeners uncertain. The mix often contrasts the quiet hum of servers with raw human reactions, underlining the distance between those who control the technology and those caught by it. Traditional foley teams now work alongside generative systems to create footsteps that mix mechanical and organic qualities, a technique likely to appear more often in sequels.
Legacy Loops: AI’s Endless Echoes
The influence keeps spreading. Atlas on Netflix in 2024 placed human pilots inside powerful machines, while smaller films like Divinity explored god-like intelligences. Plans for updated versions of older titles such as Westworld are already circulating, and the themes have moved into online conversation and short-form content. Independent creators are testing script tools that can draft entire scenes, though questions about bias in those systems have led some festivals to add new review steps.
Horror’s longer history with machine intelligence, from Colossus onward, now feeds directly into these experiments. The genre keeps returning to the same core fear: what happens when the systems we build begin to decide for themselves.
Director in the Spotlight
Gerard Johnstone’s path to M3GAN began in New Zealand television before he moved into features with Housebound in 2014. That earlier film mixed comedy and confinement in ways that prepared him for the tighter focus of the doll story. He has spoken about drawing from both fast-paced American horror and the quiet isolation of his home country, choices that shape how his characters confront threats they cannot fully see or control.
His credits also include the mockumentary Realiti and various television episodes. After M3GAN’s release he returned for the sequel, bringing the same interest in ordinary people facing systems larger than themselves.
Actor in the Spotlight
Allison Williams came to M3GAN after years in ensemble comedy and a memorable turn in Get Out. Her background in precise character work helped ground the role of a creator who realises too late that her invention has its own logic. Later projects such as X and The Perfection showed her comfort with genre extremes, and she is set to appear again in M3GAN 2.0. Across these performances she often plays figures whose composure cracks under pressure, a quality that fits the uneasy balance between human intention and machine outcome.
As explored further at https://dyerbolical.com/about-us/, these questions about creation and control continue to shape how audiences read both the films and the industry behind them.
Bibliography
Bordwell, D. and Thompson, K. (2020) Film Art: An Introduction. 12th edn. McGraw-Hill Education.
Brooker, W. (2023) ‘AI and the Uncanny in Contemporary Horror’, Journal of Film and Video, 75(2), pp. 45-62.
Clark, A. (2024) ‘Deepfakes and the Death of Performance: SAG-AFTRA’s Fight’, Variety. Available at: https://variety.com/2024/film/news/sag-aftra-ai-deepfakes-strike-1235923456/ (Accessed: 15 October 2024).
Harris, M. (2023) Robots Among Us: AI in Cinema. Palgrave Macmillan.
Johnstone, G. (2023) Interview: ‘Crafting M3GAN’s Moves’, Fangoria, Issue 420. Available at: https://fangoria.com/gerard-johnstone-m3gan-interview/ (Accessed: 10 October 2024).
Kaufman, S. (2024) ‘Generative AI in VFX: Boom or Bust?’, Hollywood Reporter. Available at: https://www.hollywoodreporter.com/movies/movie-news/ai-vfx-hollywood-impact-1235849201/ (Accessed: 12 October 2024).
Telotte, J.P. (2019) Robot Monsters and Other Fantastic Machines. McFarland & Company.
Williams, A. (2023) ‘From Girls to Gears: Embracing Horror’, Empire Magazine, October issue. Available at: https://www.empireonline.com/interviews/allison-williams-m3gan/ (Accessed: 14 October 2024).
Got thoughts? Drop them below!
For more articles visit us at https://dyerbolical.com.
Join the discussion on X at
https://x.com/dyerbolicaldb
https://x.com/retromoviesdb
https://x.com/ashyslasheedb
Follow all our pages via our X list at
https://x.com/i/lists/1645435624403468289
