LOS ANGELES, Nov 16 — The once-far-fetched idea that artificial intelligence could write, shoot and “perform” an entire feature film is no longer science fiction.

The Hollywood Reporter reported that with OpenAI’s Sora 2 generating hyper-real video from text prompts and digital “actors” such as Tilly Norwood seeking talent representation, Hollywood now feels one software update away from either a breakthrough — or an existential crisis.

For Guy Danella, president of film at indie outfit XYZ Films — best known for titles like The Raid and Skylines — the reality hit when a team pitched him a fully AI-generated 90-minute movie.

“I call it Skynet cinema because it feels like we’re coming for the humans,” he said.

“Someone asked if we’d finance it for a couple of million dollars. What happens if we say yes? What are the implications?”

The report further stated that Bryn Mooser, founder of nonfiction studio XTR, agrees the independent sector is approaching a judgment day of sorts, but insists it doesn’t need to be a Terminator-style apocalypse.

Through XTR, he has launched AI animation arm Asteria, developing projects with talents including Natasha Lyonne and Toy Story 4 writer Will McCormack.

Mooser argues that two rapidly advancing technologies are driving an industry-wide shift: increasingly powerful Nvidia chips enabling near real-time rendering, and custom-trained AI models that function as extensions of an artist’s workflow — producing storyboards, previs, backgrounds and animatics in a fraction of the time.

For major studios, he says, the combination will deliver faster and cheaper productions. For smaller creators, it could open the door to previously impossible projects.

“This shouldn’t be about making something like Anora cheaper,” he said.

“It’s about helping indie filmmakers make bigger films on the same budgets. AI has the potential to democratise studio-level storytelling.”

He notes that an indie-financed US$80 million (RM––) animated feature is “a non-starter”, but with AI tools reducing labour- and render-heavy costs, a US$10 million version could become viable.

Danella agrees that the conversation is more pragmatic than apocalyptic. “Is there a way to use AI to save enough so we can add another day of shooting? That’s the question that makes sense,” he said.

Asteria’s approach hinges on creating bespoke AI models trained on licensed or original material — explicitly avoiding the opt-out data-scraping model associated with systems like ChatGPT or Sora 2. The issue has triggered escalating warnings from rights holders that unlicensed use of copyrighted works could breach long-standing IP protections.

That concern escalated this week when a Munich court ruled that ChatGPT violated German copyright law by reproducing lyrics from nine popular songs — a landmark victory for rights society GEMA and a potential bellwether for European legal battles. OpenAI is appealing the decision.

Contracts are already shifting.

The report also further stated that at the recent American Film Market, international buyers and sales agents said disclosure clauses about AI use — covering when, where and on what datasets — are now appearing in representation and warranty language. One seller described it as “insurance against future legislation”.

The rising paper trail underscores a broader fear: that the real fight over AI will be legal, not creative.

“Our whole business is predicated on IP. If you do not protect that IP, there is no more business,” said Darren Frankel, who oversees AI initiatives at Adobe. The debate, he says, should not be “for or against” AI, but about the difference between ethical and unregulated systems. And waiting for government intervention, he cautioned, could take years.

Frankel pointed to the staggering scale of tech spending. “Amazon, Meta, Microsoft and Google are projected to spend US$364 billion this year alone, most of it on AI infrastructure. Look at the size of our industry. Where do we fit? You’re not going to hold that tide back. So how do you fight the good fight?”

While Mooser believes AI-driven disruption is inevitable, he cautioned that its full impact has yet to hit high-labour sectors like VFX and animation. “Technicolor shut down before AI even made its mark,” he said. “The question is whether you cling to the old ways, or fight to build new opportunities for the next generation.”

For all the upheaval, Danella remains committed to human-centred filmmaking.

“I believe in the flaws, the good and bad that come with human filmmaking,” he said. “The hope is to find synergy — to make better movies with more humans working more days by embracing AI as one component.”

Frankel put it bluntly: “If you don’t have that humanity, it’ll be all frosting and no cake.”