Few industries are grappling with AI’s implications as visibly—or as contentiously—as film and television. The technology is already reshaping how content is made—compressing preproduction timelines, automating postproduction workflows, and enabling shows to be green-lit that would not have existed a year ago.
In this video Explainer, McKinsey Partner Jamie Vickers and Associate Partner Alec Wrubel trace AI’s impact across the full production pipeline, from the writers’ room to distribution. They examine which players are best positioned to capture value, what risks keep industry leaders up at night, and what the landscape could look like by the end of the decade.
Both point to something the headlines tend to miss. The loudest conversation around AI in film and TV is about generated video—synthetic actors, algorithmically produced scenes. The more consequential story, they argue, is harder to see: a transformation of the entire production process, and the emergence of content formats and distribution models the industry hasn’t yet learned to name.
This interview has been edited for length and clarity.
Which aspects of AI’s impact on film and TV are people underestimating?
Alec Wrubel: Much of the conversation today has focused on AI’s impact on cost and productivity. And there is a lot to unpack and understand about those things across the value chain in the industry.
But the more interesting question, the one that often gets missed, is: How might AI completely reshape the industry in terms of structures and of fundamentally changing what content is produced and what formats that content could take? The internet, for example, didn’t just give rise to TV and film streaming. It also gave rise to open platforms like YouTube and entirely new short-form content types that we may never have anticipated.
So, of course, there will be cost and productivity impacts in how we define TV and film, and in the workflows we use today. But the more exciting prospect is unlocking entirely new content types, formats, and modes of distribution that we haven’t even fully conceptualized yet.
Jamie Vickers: When people imagine AI disrupting film and TV production, they naturally think of AI generating the video instead of humans pointing cameras at humans. And that may be a part of it over time. We’re starting to see interesting signs of that.
But that misses many of the other ways that AI is starting to show up in film and TV production. One example is the process of rush logging. You’ve done a full day of shooting, you’ve got a hundred hours of footage, and you need to apply all the metadata to that footage. That’s something that historically has involved humans doing manual review, typing in all the metadata. But that can almost certainly be automated.
Other examples would be doing reshoots in postproduction, doing much richer visualization of shoots in preproduction, or doing things like VFX [visual effects]. So, what people most often miss is equating AI in film and TV with AI-generated video. That’s actually a small slice of a much larger set of things AI can do.
Where in the production pipeline is AI having the most impact today?
Jamie Vickers: It’s almost hard to say whether AI is having the most impact in preproduction or post, because at least the most sophisticated uses I’ve seen are really rethinking the whole production process in an AI-first way.
That means, for instance, when you write the script, you’re thinking, “We’re doing this in an AI-first way. Maybe there’s a whole scene that just wouldn’t have been possible without AI.” It also translates into how you then get into the script breakdown—what has always been a largely manual process with lots of hand-drawn storyboards can now be much more automated, and the iteration cycle is so much faster.
So, I think the most sophisticated uses are not limited to scripts or to an aspect of postproduction. They’re rethinking this whole end-to-end process and asking, “How does this look different if we can weave AI through the journey, where it’s genuinely helpful or additive to do so?”
Alec Wrubel: As we think about the production process—from development through preproduction, production, postproduction, and distribution—most of the cost and productivity impact today is happening in preproduction, because it doesn’t directly affect the pixels that end up on screen.
Producers are using AI for visualization, storyboarding, script breakdowns, budgeting, shot planning, and A/B testing. It’s about being able to align the team faster and do more of the work in preproduction, so that when teams get on set, they can be more effective. In some of those select preproduction use cases, producers are telling us they’re already seeing 5 to 10 percent productivity improvements.
The other area I’d highlight is postproduction. There are more limited use cases here, in areas like dubbing, subtitling, and localization of content into different territories—so that it sounds like the actual actors, their faces moving with the dialogue—as well as match cutting: searching through hours of footage to match two shots together and edit them seamlessly.
Does faster production mean better content—and how does speed impact the creative process?
Alec Wrubel: Faster definitely does not always mean better in the production world. But there are also many places where productivity improvements will result in reinvestment rather than savings, which can enable greater creativity as well.
Think about Star Wars: A New Hope, in the 1970s, with a budget of $60 million adjusted for inflation, compared with The Rise of Skywalker [in 2019], which cost around $600 million. When CGI first came out, people might have expected: “We’ll now be able to do the same thing for much less.” Instead, what you saw was reinvestment in something with much higher visual fidelity, much more engaging visuals, at what turned out to be a much higher cost. Yes, some of this may result in doing things faster or cheaper. But some of it may also be a reinvestment in better content.
Looking for direct answers to other complex questions?
Jamie Vickers: There’s a tension around speed, and there’s a tension around quality. If you look on YouTube or TikTok, the amount of AI slop—low-quality AI-generated video—out there is enormous, despite their efforts to control it, and I’m sure that volume will continue to explode. Clearly, faster does not necessarily mean better.
The optimistic part of me wants to believe that in skillful hands, this is another enormously powerful tool for creative expression—in the same way CGI was, in the same way digital photography was. But I’m sure we will also see widespread bad use of AI tools.
What remains uniquely human in storytelling as AI becomes a creative collaborator?
Jamie Vickers: Almost every major technological innovation in media has produced a format that nobody envisioned. The camera—the original use cases were academic and scientific documentation. Nobody expected it to become a major new art form. Then there was the introduction of Instagram, social media, and massive disruptions to what it means to express yourself creatively.
Alec Wrubel: There’s a bear view and a bull view. The bear view is what we’ve seen with Sora—a huge volume of AI-generated content that some might call slop, and a model that couldn’t support a viable business. The bull view is that AI expands what’s possible for authentic human-led storytelling and increases the value and the premium of that storytelling.
What cuts through the noise is authentic, human-led content—the stories people connect with, where taste plays a role. I’m excited about the potential for independent studios or undiscovered producers to make the kinds of things that would previously have been unthinkable, that would require big budgets and major studio support. I’m hopeful that high-quality human storytelling will always cut through the noise, no matter how much content there is.
Does AI democratize filmmaking or concentrate power further?
Jamie Vickers: Most big technological innovations in the past required enormous up-front investment just to use. Take digital filmmaking: to get it right, you had to buy a digital camera, which could cost hundreds of thousands or even millions of dollars. You had to have the software and the employees able to work with it. CGI is an even starker example: Most of the biggest-budget films of all time were made since the emergence of CGI.
AI is different. With AI, anyone who can get a subscription—which currently costs low double-digit dollars a month—can start experimenting and creating content. That’s a fundamental difference from past examples of technological disruption, and it gives good reason for optimism that this will democratize at least certain aspects of storytelling.
Alec Wrubel: The honest answer is that AI probably continues to democratize filmmaking and to concentrate power. Over the past few years, seven distributors in the US have made up something like 85 percent of total US original content spend—that’s concentration. But at the same time, we’ve seen massive democratization on YouTube and TikTok, with creators like MrBeast producing premium, hour-long content that approximates a TV show.
More productivity and less expensive content don’t erase the advantages of distribution. Those major studios still have massive marketing capability, recognizable IP, and access to A-list talent. But I think niches will form. You may see a hollowing out of the middle, as the market concentrates in top-end premium storytelling and, at the lower-end, a long tail of grassroots creators now able to make higher-quality content.
Who stands to gain—and who is most at risk?
Alec Wrubel: The players who stand to gain the most are distributors, who will continue to be the choke point for where content is released. Some tech players may capture value if their models aren’t commoditized and they’re able to build distinctive tools used by industry incumbents.
Those who might be most at risk are players across the value chain who don’t think through what AI means for their business model. Concern about the middle is valid—the middle of studios, streaming services, and creative talent.
Jamie Vickers: Distributors should be well-placed. Five players globally command a majority of content spend, while the studio side is much more fragmented—hundreds and thousands of studios doing content production, including many quite small companies. The tech vendors selling this technology are clearly well-positioned to benefit. How much of that benefit they capture is an interesting and still unresolved question.
But perhaps the group best placed to benefit is consumers. Ultimately, this should translate into more content choices at a better price; at least I hope that will be one of the things that comes out of it. And I think the theory supports it.
What are the biggest risks industry leaders are wrestling with right now?
Jamie Vickers: There is clearly a question around end users’ acceptance of AI-generated content. Historically, it takes people a while to get used to new technology—but in the end, they do. At the moment, you see a lot of audience rejection when they know content is AI-generated.
What AI could mean for film and TV production and the industry’s future
The second risk is rejection by creative communities. We’ve seen this with SAG-AFTRA’s [labor union strike] and many other examples. If adopting AI comes at the expense of the storytellers and creative talent that make this industry magical, that’s not a step worth taking.
The third risk—and it’s more of a barrier to adoption—is simply, what is it going to look like to use this technology at scale? New features in existing software? Custom in-house builds? That whole question of change management and commercial rollout is still very much unresolved.
Alec Wrubel: Leaders across the industry consistently raise three risks. First: talent and creative implications—very visible in the SAG-AFTRA and Writers Guild strikes, and in the debate around The Brutalist, which used AI dubbing in the creation of some accents. Digital likeness also comes up here: talent agencies are working to protect and license how their clients’ likenesses are used.
Second: IP and rights infringement. Some models are trained on all content on the internet, some of which is copyright protected—that’s a valid concern.
Third: hallucination and bias. In casting workflows, marketing, script development—is the model accurately representing people and diversity of opinion, or is it biased toward one type of person?
What does the film and TV industry look like by 2030?
Alec Wrubel: There are three scenarios in descending order of likelihood. First: point solutions that change current production workflows—innovations in pre- and postproduction expanding to physical production and AI-assisted improvements across the process. Our analysis indicates that by 2030, $10 billion of US original content spend could be addressed by some form of AI.
Second: wide-scale democratization of professional-grade content creation—the same way YouTube democratized video distribution, AI could make pro-level content capabilities accessible to an even wider range of creators.
Third, and potentially most interesting: the creation of entirely new content formats and distribution channels. When Netflix started its DVD business, few people predicted it would become one of the largest producers of original content in the world. If adoption plays out as we expect, there is potentially $60 billion of revenue at stake that could be redistributed over the next five years.
Jamie Vickers: The future is already showing up in the present. In 2026, we are seeing shows get green-lit that likely would not have happened without AI. That’s not a prediction—it’s already showing up in financial numbers for some studios and broadcasters this year.
Looking further out, forward-learning studios that get this right will have the opportunity to make equivalent content for around 30 percent less effort. Whether they choose to reinvest that in more spectacular productions or use it to make more films and shows is an open question. But if you believe this is a major productivity lever, then the right thing to do as a company is to lean in and adopt quickly—because years one, two, and three are your window to capture market share before everyone else catches up.
