AI in movies and animation
In March 2023 an AI clip of Will Smith eating spaghetti was a joke. Three years later the same test came out close to the real thing, studios were suing AI companies and signing with them, and the unions had written AI into their contracts.
Every entry dated and sourced · updated
Watch
DREAMS OF VIOLETS | Official Announcment Trailer (2026) · FOUNTAIN 0 studios Dreams of Violets, the first all-AI feature at a major festival June 2026
Veo 3.1 and more artistic control in Flow · Google DeepMind Google releases Veo 3.1 October 2025
Introducing Sora 2 · OpenAI OpenAI launches Sora 2 and the Sora app September 2025
ANCESTRA Trailer · Primordial Soup Ancestra premieres at Tribeca June 2025
2026
Runway’s fourth AI Festival
Runway’s festival, now called AIF, held galas in New York, Los Angeles and Tokyo with tracks for film, design, new media, fashion, advertising and gaming. Films had to run 3 to 15 minutes and use generated video, and the top prize was $50,000 plus Runway credits. Jurors included Roger Avary and Tribeca’s Jane Rosenthal.
“Welcome to Runway's fourth annual AI Festival.” aif.runwayml.com
Dreams of Violets, the first all-AI feature at a major festival
Dreams of Violets, a feature about the January 2026 crackdown in Iran made entirely with AI by Ash and Pooya Koosha, premiered at the Tribeca Festival. Variety called it the first fully AI-generated film programmed at a major festival, found it dramatically numbing, and noted its budget was $2,000.
““Dreams of Violets,” which premiered last week at the Tribeca Festival, is the first movie generated entirely by AI to be programmed at a major film festival — and it’s also the first movie generated entirely by AI that I’ve seen.” variety.com
DREAMS OF VIOLETS | Official Announcment Trailer (2026) · FOUNTAIN 0 studios Lionsgate takes a stake in Runway and plans an AI series from its own IP
Lionsgate expanded its Runway partnership, took an equity interest in the company and set up a joint program to develop new projects. The first is a short-form episodic series drawing on existing Lionsgate properties and made with Runway’s generative models.
“Under the new agreement, Lionsgate has taken an equity interest in Runway and the two companies will be launching a joint development program to develop and produce new IP.” runway.com
Actors win limits on synthetic performers
SAG-AFTRA members ratified a four-year contract with 91 percent in favor. It builds on the 2023 replica rules and says producers may use AI-made synthetic performers instead of human actors only when they add significant value.
“With the contract, producers are only allowed to use AI performers if they bring “significant additional value” compared to a live actor or that actor’s digital avatar.” lamag.com
Critterz misses its Cannes premiere
Critterz, the OpenAI-backed animated feature meant to be the first mainstream film made through a generative AI pipeline, missed its planned Cannes premiere after OpenAI shut down Sora. Its producers showed early footage to buyers in the Cannes market instead.
“The highly anticipated animated feature Critterz, which was to be the first mainstream commercial film created through a generative AI pipeline, failed to secure its planned in-festival premiere at Cannes this week.” thenews.com.pk
Astana AI Film Festival opens its first call
Kazakhstan’s first AI film festival opened a worldwide call for short films made with AI, up to ten minutes long. Entries ran from May through August 2026 for a festival in the autumn, open to teams and individuals.
““According to the competition rules, participants must create an AI-generated short film of up to 10 minutes and upload it to the festival’s platform along with a project description.” astanatimes.com
OpenAI shuts down Sora
OpenAI closed the Sora app and website on April 26, 2026, and the Sora API on September 24, 2026. Sora 2 and its app had launched only the previous September, and the closing left projects built on it looking for other tools.
“The Sora web and app experiences were discontinued on April 26, 2026.” help.openai.com
Writers' new contract covers licensing scripts to train AI
Writers ratified a four-year contract that keeps the 2023 AI rules and adds terms for when studios license writers' scripts to train generative AI. Studios must notify the guild, which can then ask to discuss the license and pay for the writers.
“In terms of AI protections, the 2026 MBA preserves the protections from 2023 and adds requirements if the studios seek to license writers’ scripts or projects based on those scripts that get produced to Generative Artificial Intelligence (GAI) systems.” screendaily.com
ByteDance launches Seedance 2.0
ByteDance launched Seedance 2.0, which makes clips of up to 15 seconds from a text prompt. A clip of Tom Cruise fighting Brad Pitt, made from a two-line prompt, spread on X within the week and drew condemnation from the studios and SAG-AFTRA.
“Similar to tools such as OpenAI’s Sora, Seedance allows users to create videos (currently limited to 15 seconds in length) by just entering a text prompt.” techcrunch.com
2025
Disney licenses its characters to OpenAI’s Sora and invests $1 billion
Disney signed a three-year license letting Sora users make short videos with more than 200 Disney, Marvel, Pixar and Star Wars characters, and agreed to a $1 billion equity investment in OpenAI. A selection of the fan-made videos is to stream on Disney+. Six months earlier Disney had sued Midjourney over the same kind of use.
“The Walt Disney Company and OpenAI have reached an agreement for Disney to become the first major content licensing partner on Sora, OpenAI’s short-form generative AI video platform, bringing these leaders in creativity and innovation together to unlock new possibilities in imaginative storytelling.” thewaltdisneycompany.com
Getty Images loses its main UK case against Stability AI
Getty sued Stability AI in 2023, saying millions of its photos were used to train Stable Diffusion without permission. The UK High Court sided mostly with Stability, finding for Getty only on limited claims tied to its watermark showing up in generated images.
“As reported by The Guardian, Stability AI prevailed in a high-profile UK High Court case, following Getty first suing the company in 2023 for allegedly using its copyright images to train its Stable Diffusion AI art tool without permission.” engadget.com
Google releases Veo 3.1
Google released Veo 3.1 with richer sound, more narrative control and more realistic textures, and added sound to Flow’s editing tools. Google said Flow had generated more than 275 million videos in its first five months.
“We’re also introducing Veo 3.1, which brings richer audio, more narrative control, and enhanced realism that captures true-to-life textures.” blog.google
Veo 3.1 and more artistic control in Flow · Google DeepMind OpenAI backs Critterz, an animated feature made with its tools
An animated feature called Critterz, based on a 2023 AI short by OpenAI’s Chad Nelson, was announced with OpenAI’s backing. Its makers aimed to finish it in nine months for under $30 million and premiere at Cannes in 2026, hiring artists to design the characters and actors to voice them.
“The goal of Critterz is to make the movie in nine months on a budget of less than $30 million, which is exponentially shorter and cheaper than an animated movie usually takes to make.” avclub.com
OpenAI launches Sora 2 and the Sora app
OpenAI released Sora 2, which generates dialogue and sound effects in sync with the picture and obeys physics better than before. It came with a social app where people could put themselves and their friends into scenes.
“It also features synchronized dialogue and sound effects.” openai.com
Introducing Sora 2 · OpenAI Disney, NBCUniversal and Warner Bros. Discovery sue China’s MiniMax
Three studios jointly sued the Chinese AI company MiniMax over its Hailuo service, which they say generates images and video of their characters on demand. It was the studios' third suit against an AI company in four months.
“The studios allege MiniMax’s Hailuo AI service generates unauthorised images and videos of iconic characters, describing it as a “Hollywood studio in your pocket” built on stolen intellectual property.” irishlegal.com
AI actress Tilly Norwood draws Hollywood’s anger
Producer Eline Van der Velden introduced Tilly Norwood, an AI-generated actress, at the Zurich Film Festival and went looking for an agent to represent her. SAG-AFTRA answered that Norwood is not an actor and was trained on real performers' work without permission or pay.
““To be clear, ‘Tilly Norwood’ is not an actor, it’s a character generated by a computer program that was trained on the work of countless professional performers — without permission or compensation,” SAG-AFTRA wrote.” techcrunch.com
Warner Bros. Discovery sues Midjourney
Warner Bros. Discovery became the third major studio to sue Midjourney, over images and video of Superman, Batman, Bugs Bunny and Scooby-Doo. Its complaint set Midjourney’s outputs beside stills from its own films and said the company could block such prompts but chose not to.
“"It is hard to imagine copyright infringement that is any more willful than what Midjourney is doing here," the plaintiff added.” engadget.com
Runway festival films play in IMAX theaters
The ten winning films of Runway’s 2025 AI Film Festival were given special limited screenings at ten AMC cinemas in American cities in August 2025, nine of them IMAX theaters. It put AI-made shorts on commercial cinema screens.
“The AIFF was held as an 'IMAX Presents' event, and the 10 winning films were scheduled to be screened as 'special limited screenings' in theaters equipped with the IMAX system.” gigazine.net
Netflix publishes rules for partners using generative AI
Netflix posted guidance telling its production partners when they may use generative AI freely and when they need written approval, such as for final deliverables or an actor’s likeness. It came after a Netflix documentary drew criticism for apparently AI-made archival photos.
“This week, Netflix published a post on its Partner Help Center hub detailing its guiding principles regarding gen AI and the situations in which it is ok with production teams using the technology.” theverge.com
Netflix puts generative AI footage on screen in El Eternauta
Netflix co-CEO Ted Sarandos said a building collapse in the Argentine series El Eternauta was made with generative AI, the first such final footage in a Netflix original. He said the shot was finished ten times faster than with traditional effects work.
“The executive says that The Eternaut features "the very first Gen AI final footage to appear on screen in a Netflix original series or film." Clearly, the show is also a prototype for how Netflix can avoid costs it doesn't want to swallow in the future.” engadget.com
AMC Networks brings Runway into marketing and TV development
AMC Networks, home of Mad Men and Breaking Bad, became the first cable company to sign with Runway. It plans to use Runway’s models for promotional clips, campaign ideas and pre-visualizing shows still in development.
“Runway and AMC Networks, the international entertainment company known for popular and award-winning titles including MAD MEN, BREAKING BAD, BETTER CALL SAUL, THE WALKING DEAD and ANNE RICE’S INTERVIEW WITH THE VAMPIRE, are partnering to incorporate Runway’s AI models and tools in AMC Networks’…” runway.com
Ancestra premieres at Tribeca
Ancestra, a short by Eliza McNitt made with Google DeepMind’s Veo and produced by Darren Aronofsky’s Primordial Soup, premiered at the Tribeca Festival. It was the first of three shorts the partnership planned to make with Veo.
“Eliza McNitt’s immersive short film ANCESTRA, executive produced by fellow Sloan grantee Darren Aronofsky made its world premiere at the 2025 Tribeca Festival on June 13.” scienceandfilm.org
ANCESTRA Trailer · Primordial Soup Disney and Universal sue Midjourney
Disney and Universal jointly sued Midjourney for copyright infringement, the first time major studios took an AI image company to court. The complaint shows how easily Midjourney recreates their characters and says Disney’s earlier letter asking it to stop got no real answer.
“Earlier today, Disney and Universal teamed up to file a lawsuit alleging copyright infringement against Midjourney, the creator of the Midjourney AI system.” plagiarismtoday.com
Midjourney releases its first video model
Midjourney, the leading image generator, released V1, its first video model. Users press Animate to turn an image into four five-second clips, at roughly the cost of one image per second of video.
“This means that you still make images in Midjourney, as normal, but now you can press “Animate” to make them move.” updates.midjourney.com
Introducing Midjourney V1 Video · Midjourney Total Pixel Space wins Runway’s festival at Lincoln Center
Runway’s third AI Film Festival opened at Lincoln Center’s Alice Tully Hall, and Jacob Adler’s Total Pixel Space took the top prize. Runway said submissions had grown from about 300 in the first year to about 6,000.
“The one and half-hour lineup stretched across a range of creative styles and ambitious themes — with Jacob Adler's “Total Pixel Space " taking home the festival's top prize.” clickondetroit.com
Total Pixel Space · Jacob Adler An AI Darth Vader voice in Fortnite draws a union charge
Fortnite added a Darth Vader whose voice was generated by AI. SAG-AFTRA filed an unfair labor practice charge against Epic Games' subsidiary Llama Productions, saying the AI voice replaced its members' work without notice to the union or a chance to bargain.
“LOS ANGELES – Hollywood's actors' union filed an unfair labor practice charge against Llama Productions on Monday, alleging the company replaced actors' work by using artificial intelligence to generate Darth Vader's voice in Fortnite without notice.” clickondetroit.com
Google’s Veo 3 adds sound
Google released Veo 3, the first of its video models to generate sound with the picture: background noise, birdsong and dialogue with lip sync. It launched with Flow, a filmmaking tool built around the model.
“Veo 3, our new state-of-the-art video generation model, not only improves on the quality of Veo 2, but for the first time, can also generate videos with audio — traffic noises in the background of a city street scene, birds singing in a park, even dialogue between characters.” blog.google
Introducing Veo 3: Old Sailor · Google Prime Video starts AI-aided dubbing
Prime Video began a pilot of AI-aided dubbing in English and Latin American Spanish on 12 licensed movies and series that had no dub. Amazon said human localization professionals check the AI’s work.
“Starting today, AI-aided dubbing in English and Latin American Spanish will be available initially on 12 licensed movies and series, including titles such as El Cid: La Leyenda, Mi Mamá Lora, and Long Lost.” aboutamazon.com
Runway releases Gen-4
Runway released Gen-4, which keeps the same character, place or object consistent from shot to shot using a single reference image, with no extra training. That is what lets separate generated shots be cut together into one scene.
“Runway says that Gen-4 allows users to generate consistent characters across lighting conditions using a reference image of those characters.” techcrunch.com
Introducing Runway Gen-4 | Runway · Runway The Brutalist’s editor reveals AI voice work during Oscar season
The editor of The Brutalist said the film used Respeecher’s AI to refine Adrien Brody and Felicity Jones’s Hungarian pronunciation, and generative AI for drawings in the epilogue. The disclosure set off a backlash while the film was an awards favorite.
“The filmmakers behind The Brutalist, a likely Oscar contender currently being distributed by A24, used AI to alter actor's dialogue and create images used in the film's epilogue, the film's editor Dávid Jancsó shared in an interview with RedShark News.” engadget.com
2024
The Animation Guild ratifies a contract with AI terms
Animation Guild members approved their 2024 to 2027 contract with 76 percent of the vote. Its AI terms include the right to consult with productions on alternatives to generative AI tools, but some members of the negotiating committee voted no over the AI provisions.
“Per the guild, 76.1 percent of voting members chose to accept the deal, which covers 2024-27 and includes wage increases and key protections for members with regard to use of artificial intelligence in animation.” hollywoodreporter.com
Google releases Veo 2
Google released Veo 2, its second video model. It understood the language of the camera, so a prompt could ask for a lens, a tracking shot or shallow depth of field, at resolutions up to 4K.
“Veo 2 understands the unique language of cinematography: ask it for a genre, specify a lens, suggest cinematic effects and Veo 2 will deliver — at resolutions up to 4K, and extended to minutes in length.” blog.google
Creating with Veo 2 · Google DeepMind Lionsgate signs with Runway for a model trained on its catalog
Lionsgate became the first Hollywood studio to partner with the AI video company Runway. Runway agreed to build a custom video model trained on Lionsgate’s own film and television catalog, for its filmmakers to use in pre-production and post-production.
“Lionsgate and Runway have entered into a first-of-its-kind partnership centered around the creation and training of a new AI model, customized on Lionsgate’s proprietary catalog.” runway.com
Video game actors strike over AI
SAG-AFTRA’s voice and motion-capture performers struck the major game companies over AI protections. The strike ran nearly a year and ended in July 2025 with a contract requiring consent and disclosure for AI digital replicas, approved by 95 percent of voters.
“The strike was authorized in September 2023 and started after a year and a half of negotiations which failed to result in a protection agreement from the use of artificial intelligence (AI) for all performers covered by the interactive media agreement.” en.wikipedia.org
Kuaishou releases Kling
Kuaishou, the Chinese short-video company, released Kling, a model it said could make clips up to two minutes long in 1080p. Unlike Sora, people could try it right away, and it showed a Chinese company competing directly with OpenAI on video.
“But unlike Sora, which still remains inaccessible to the public four months after OpenAI trialed it, Kling soon started letting people try the model themselves.” technologyreview.com
Runway releases Gen-3 Alpha
Runway released Gen-3 Alpha, the first model trained on its new large-scale infrastructure. It brought a clear jump in image quality, consistency and motion over Gen-2, and Runway framed it as a step toward models that simulate the world.
“It is a major improvement in AI model fidelity, consistency and motion over Gen-2, and a step towards building General World Models.” runwayml.com
Gen-3 Alpha: Available Now | Runway · Runway OpenAI shows Sora
OpenAI showed Sora, a text-to-video model that made photorealistic clips up to a minute long, far longer and steadier than anything public at the time. At first only red teamers testing it for harms could use it.
“The AI company says Sora “can create realistic and imaginative scenes from text instructions.” The text-to-video model allows users to create photorealistic videos up to a minute long — all based on prompts they’ve written.” theverge.com
Introducing Sora — OpenAI’s text-to-video model · OpenAI
2023
Actors ratify the first contract with AI replica rules
SAG-AFTRA members ratified the deal that ended their 118-day strike, with 78 percent in favor. It requires consent and pay before a studio makes or uses a digital replica of a performer, though many members said the AI terms did not go far enough.
“The deal provides a 7% increase in minimum rates in the first year and the first-ever protections against the use of artificial intelligence to replicate performances, requiring compensation and consent guardrails to protect performers from generative AI technology.” abc7chicago.com
The writers' strike ends with the first AI rules in a Hollywood contract
The Writers Guild’s new contract says AI cannot write or rewrite scripts and AI output cannot count as source material. Studios must also tell writers when material handed to them was made by AI.
“AI can’t write or rewrite literary material, and AI-generated material will not be considered source material under the MBA, meaning that AI-generated material can’t be used to undermine a writer’s credit or separated rights.” wgacontract2023.org
Marvel’s Secret Invasion opens with AI-made credits
Marvel’s Disney+ series Secret Invasion debuted with an opening title sequence its director confirmed was made with AI by Method Studios. Artists objected, and Method said no artists' jobs were replaced.
“Marvel’s latest series, “Secret Invasion,” made its debut on Disney+ today, sparking tons of backlash after it was confirmed the intro sequence was AI-generated.” techcrunch.com
Runway announces Gen-2, text to video
Runway announced Gen-2, a model that makes short new video clips from a line of text, an image or another clip. The Verge called the output rudimentary, and Runway said users would get access within weeks.
“Runway says it’ll be making its new generative video model, Gen-2, available to users in ‘the coming weeks.’” theverge.com
Introducing Gen-2: Text to Video | Runway · Runway Runway holds the first AI Film Festival
Runway screened the first festival for films made with AI in New York in February 2023, then in San Francisco in March. Hundreds of films were submitted and ten finalists were shown. The Grand Prix went to Generation by Riccardo Fusetti.
“Introducing the first annual AI Film Festival - powered by Runway.” aif.runwayml.com