Hollywood’s AI Experiment Is Getting Much Bigger

Hollywood’s artificial intelligence experiment is no longer happening somewhere off in the distance. It is already moving through the entertainment business, appearing in production workflows, visual effects, pre-production, post-production, development and the business decisions happening behind the scenes. What is changing now is not simply the technology itself, but the scale at which the industry is beginning to test it.

The conversation has also become more complicated. Hollywood is no longer debating only whether artificial intelligence belongs in entertainment. Studios, filmmakers and technology companies are increasingly dealing with a more practical question: where should AI fit, and where should human creative control remain essential?

That distinction is becoming particularly visible as September comes to a close. New deals, production tools, studio experiments and technology partnerships are appearing alongside continuing debates over copyright, likeness, employment and creative ownership. Axios reported earlier this month that Hollywood studios are beginning to strike new AI-related deals as adoption grows, suggesting that the industry is moving from broad arguments about the technology toward more concrete arrangements for using it.

That shift is significant because Hollywood has spent the past several years treating AI as both an opportunity and a threat.

Now it is increasingly becoming part of the infrastructure of entertainment.

The easiest mistake is to imagine Hollywood’s AI experiment as a push toward entirely computer-generated movies. That is not where most of the industry’s current activity is concentrated. Instead, AI is being introduced into individual stages of a production.

A filmmaker might use it to explore visual concepts before filming begins. A production team might use AI-assisted software to organize a script or identify elements that need to be prepared for a shoot. Visual-effects artists can experiment with generated assets. Editors can use AI-assisted functions during post-production. Sound teams can use technology to clean or manipulate audio. Other systems can help with translation, dubbing, scheduling and administrative work.

The technology does not have to make the entire movie to change how a movie gets made.

That may ultimately be the more important story.

A 2026 industry outlook from ProdPro found that studio executives planned to apply AI tools to an average of 32% of projects on their 2026 slates, up from 29% the previous year. The report identified visual effects as the area expected to see the greatest near-term impact, while also pointing to applications in pre-production such as script breakdowns, scheduling and budget modeling.

Those numbers illustrate how quickly the discussion is moving into ordinary production decisions.

AI does not have to replace a director, cinematographer, editor or visual-effects artist to have an effect on their work. If a production team can use software to complete a repetitive task in minutes instead of hours, that changes the economics and workflow surrounding the project.

The bigger question becomes what happens to the time that technology saves.

Does it allow artists to spend more time experimenting?

Does it allow smaller productions to attempt visual ideas that would previously have been too expensive?

Does it reduce repetitive work?

Or does it encourage studios to reduce the number of people involved?

Those questions do not have one universal answer, and that is part of what makes the current Hollywood AI conversation so complicated.

There are already signs that studios are quietly building AI capabilities while publicly continuing to negotiate its limits. The Los Angeles Times reported in July that more than one in ten of hundreds of entertainment job postings it examined in late June appeared likely to be connected to AI. The postings included jobs focused on developing AI tools as well as positions designed to protect studios’ intellectual property against unauthorized AI use.

In other words, Hollywood is preparing for both sides of the technology.

It wants to use AI.

It also wants to control AI.

And it wants to protect its own content from being used by AI systems without permission.

That tension is becoming one of the defining characteristics of the industry’s transition.

Technology companies are also increasingly interested in Hollywood’s enormous libraries of intellectual property. In August, the Los Angeles Times reported that Google had approached representatives at Disney, Universal, Warner Bros. Discovery and other studios about potentially licensing intellectual property for use with AI models. The discussions were complicated by copyright, labor and other legal concerns.

That development points toward an entirely different kind of entertainment technology business.

For years, Hollywood’s most valuable assets were films, television programs, characters and recognizable franchises.

Now those same assets can potentially become training material, licensed digital assets or inputs for new AI-assisted systems.

That creates an obvious commercial opportunity.

It also creates an enormous legal and creative question.

Who gets permission?

Who gets paid?

How much control does the original rights holder retain?

Can an actor’s likeness be used?

Can a studio’s characters be reproduced?

What happens when an AI-generated result looks too similar to existing intellectual property?

And what does authorship mean when a human gives instructions to a machine that produces part of the final result?

These questions are becoming less theoretical.

The entertainment industry has already spent years negotiating over performers’ likenesses and AI-generated replicas. The Los Angeles Times reported in June that Hollywood’s AI debate had shifted toward issues including consent, likeness and compensation, while describing new mechanisms designed to give performers greater control over how their digital identities can be licensed.

Voice performers face a particularly complicated version of the problem.

A voice is both a creative performance and a recognizable personal identity. AI cloning technology can reproduce aspects of a voice with increasing sophistication, creating potential opportunities for performers who want to license digital replicas while also creating concerns for performers who fear their voices could be used without meaningful consent or compensation.

The Los Angeles Times reported in August that voice actors were divided over AI cloning, with some established performers licensing replicas while freelancers faced concerns about lost work and unauthorized uses of their voices.

That makes the Hollywood AI conversation fundamentally different from a simple debate about software.

It is also a labor conversation.

And a rights conversation.

And an ownership conversation.

But there is another side of the experiment that is easy to overlook: AI may allow people outside the traditional Hollywood system to make things that would previously have required enormous resources.

Lower production costs can create opportunities for independent filmmakers, small studios and creators.

That possibility is one reason some entertainment executives and technology companies are enthusiastic about the technology.

Jeffrey Katzenberg, for example, announced plans in September to launch an animation studio focused on using AI in film and television production. He has argued that lower barriers and costs could lead to more productions and new forms of storytelling.

The idea is familiar from previous technological revolutions.

Digital cameras reduced some of the barriers to filmmaking.

Affordable editing software changed independent production.

Streaming changed distribution.

Social media changed promotion.

AI could change the production process itself.

But every technological revolution creates winners, losers and unexpected consequences.

One of the most interesting questions is whether AI will ultimately make Hollywood more creative or simply more efficient.

Those are not necessarily the same thing.

Efficiency is relatively easy to measure.

A task takes less time.

A production costs less.

A draft is generated faster.

A visual concept can be tested quickly.

A translation can be produced more efficiently.

But creativity is harder to measure.

A generated image may look impressive while failing to understand why a particular visual choice matters to the story.

A generated sequence may contain beautiful individual shots while struggling with continuity.

A generated character may look convincing in one frame but behave inconsistently across an entire production.

Those limitations are still very real.

Autodesk recently demonstrated exactly that problem with an AI-generated film at Autodesk University 2026. The project contained a noticeable continuity error in which a character changed position between shots, and attempts to correct the problem through prompting did not solve it. Autodesk executives used the example to emphasize that professional production requires consistency across shots, revisions and the larger creative pipeline.

That may sound like a small technical problem.

For filmmaking, it is not.

Movies depend on continuity.

Characters need to remain recognizable.

Locations need to make sense.

Objects need to stay where they are supposed to be.

Lighting needs to match.

Movement needs to remain coherent.

Story details need to connect across scenes.

An AI system that can create one spectacular image is not necessarily an AI system capable of managing an entire feature film.

That difference between generating something impressive and generating something usable is becoming increasingly important.

The entertainment industry is therefore likely to continue using AI in pieces rather than treating it as one giant replacement for the entire filmmaking process.

A visual artist may use one system.

An editor may use another.

A writer may use an AI assistant for brainstorming.

A producer may use software for planning.

A post-production team may use another tool for cleanup.

The final product may contain technology at multiple stages without ever being entirely generated by AI.

That is already reflected in the technology being demonstrated across the professional production sector.

At IBC2026 in September, industry researchers and production companies presented work involving AI-assisted visual-effects workflows, AI-supported script breakdowns, virtual-production systems and tools intended to improve production pipelines.

This is a more practical vision of AI in entertainment.

Instead of a machine replacing the studio, the machine becomes another layer inside the studio.

The bigger Hollywood question is therefore becoming one of boundaries.

What should AI do?

What should humans do?

What should remain entirely human?

And who gets to decide?

Those questions become particularly important when the technology moves from behind-the-scenes assistance into recognizable performances.

Digital humans and synthetic actors have already begun appearing in international entertainment experiments. The Los Angeles Times reported this month on China’s rapidly developing AI film industry, including the use of actors’ facial likenesses to create digital characters for AI-generated productions.

China’s experience is worth watching because it shows what can happen when AI-generated entertainment becomes part of a broader production ecosystem.

Reuters reported in September that Chinese cities and industrial parks were actively competing to attract AI filmmakers, offering incentives and building technology clusters around AI-generated film production.

Hollywood does not have to follow exactly the same path.

But it cannot ignore the experiment happening elsewhere.

The global entertainment industry is effectively running multiple versions of the AI experiment at the same time.

Hollywood is testing studio workflows and negotiating labor and intellectual-property boundaries.

Technology companies are building increasingly sophisticated generation and production systems.

Independent filmmakers are experimenting with lower-cost production.

China is developing large-scale AI film ecosystems.

Film festivals are increasingly discussing the technology.

And audiences are beginning to encounter AI-generated entertainment whether they actively seek it out or not.

That makes the next stage less about whether AI is coming.

It is already here.

The real question is what kind of entertainment industry it creates.

There is a version of the future in which AI handles tedious production tasks and gives artists more time to focus on storytelling.

There is another version in which studios use the technology primarily to reduce costs.

There is a version in which independent creators gain access to production capabilities that were previously available only to major companies.

There is another in which the internet becomes flooded with inexpensive synthetic entertainment and audiences become increasingly selective about what feels authentic.

All of those possibilities can exist at once.

And Hollywood is still figuring out where it wants to land.

That uncertainty may actually be the most interesting part of the current moment.

The industry is experimenting in public and private.

Studios are making deals.

Technology companies are courting entertainment companies.

Filmmakers are testing new workflows.

Actors and writers are negotiating protections.

Visual-effects teams are exploring new systems.

And audiences are beginning to learn how to distinguish between technology that genuinely improves a production and technology that simply produces something quickly.

The biggest AI experiment in Hollywood, therefore, may not be about making an entirely artificial movie.

It may be about redesigning the process that produces movies in the first place.

That process could eventually become faster, cheaper and more flexible.

But the technology still has to answer the oldest question in entertainment: does it help tell a better story?

That is where Hollywood’s AI experiment ultimately becomes more than a technology story.

It becomes a creative story.

The tools are getting bigger.

The investment is getting bigger.

The experiments are getting bigger.

The legal questions are getting bigger.

The labor questions are getting bigger.

And the number of people trying to figure out what comes next is growing just as quickly.