Who is on the Hollywood Reporter AI power list for entertainment?

The Hollywood Reporter's first AI power list, 'The AI 25,' features CEOs, investors, board members, and executives who control AI infrastructure in entertainment — including the CEO of Stability AI, co-founders of AI startups, and filmmakers like James Cameron, Ben Affleck, and Darren Aronofsky who have taken institutional roles. Notably absent are working filmmakers known for the craft of AI prompting itself.

Why does craft not appear on AI filmmaking power lists?

Power lists measure leverage — board seats, equity stakes, regulatory access, and deal flow — rather than creative skill. Craft in AI filmmaking means iterating on a prompt dozens of times to match a precise vision, which does not scale and does not generate the kind of institutional visibility that makes headlines. The people who are genuinely skilled at working with AI models tend to operate invisibly, in contrast to platform owners and investors who shape the conditions everyone else works within.

Will AI filmmaking power shift from tech executives to creative practitioners?

Historically, every major technology transition in filmmaking has eventually moved authority from infrastructure owners to skilled practitioners. Edison's patent monopoly gave way to the filmmakers who moved to Hollywood; studio moguls ceded ground to directors like Coppola and Scorsese. The current AI power landscape — dominated by platform owners, investors, and policy shapers — is likely a transitional phase. As tools stabilize and creative vocabulary becomes the primary differentiator, the balance of power is expected to shift toward those who can actually use the tools with precision and artistic intent.

The power list -- CinePrompt Field Notes

The Hollywood Reporter published "The AI 25" this week. First time the trade has assembled a formal power list for AI in entertainment. Twenty-five names, alphabetical, each with a headshot and a paragraph about their vision for what Hollywood should become.

Read the list carefully.

CEOs. Board members. Investors. Activists. Entrepreneurs. A chief AI officer. A former design ethicist from Google. A training-data billionaire whose primary qualification is that she is richer than Taylor Swift. A founder who told the Toronto International Film Festival that slop was not really a thing. The co-CEOs of a company that just killed its own video product.

Count the filmmakers. Cameron, who joined a board. Affleck, who secretly built a startup. Aronofsky, who runs a company called Primordial Soup. Bateman, who runs a festival devoted to not using the technology at all.

Nobody on the list is there because they made something beautiful with a prompt and a model.

What the list measures

The AI 25 measures leverage. Board seats. Equity stakes. Regulatory access. Deal flow. Studio relationships. The ability to shape the conditions under which AI filmmaking happens. Where the money flows. Where the IP goes. Where the guardrails get placed or removed.

This is useful information. Power matters. The people who control the platform, the investment, the licensing deal, and the rulebook determine the landscape that everyone else works inside.

But the list reveals its own assumptions by what it does not include. Not a single name appears because they are good at this. Not one person who sits at a prompt and works until the output matches the picture in their head. Not the twenty-two-year-old in Switzerland who used AI's dissociated quality to express dementia for five hundred euros. Not the eighty-eight-year-old in Paris who said the tools gave him his childhood back. Not the filmmakers from a hundred and thirty-nine countries who submitted work to the Higgsfield competition and proved that talent was never the barrier.

The list says: power in AI filmmaking belongs to the people who own the pipes, build the platforms, fund the studios, set the rules, and occasionally sit on a board to lend their name prestige. Power does not belong to the person in a room at two in the morning trying to get the rim light to fall on the correct side of a face that does not exist.

The infrastructure class

Hollywood has always organized its power around infrastructure. Studios owned the sound stages, the distribution networks, the financing apparatus, and the exhibition chains. The talent worked inside that infrastructure. The infrastructure set the terms.

The AI 25 is the infrastructure class for a new kind of studio system. The generation platforms are the new sound stages. The API keys are the new distribution licenses. The training data agreements are the new talent contracts. The people who control those assets appear on the list. The people who use them do not.

Prem Akkaraju, CEO of Stability AI, says something worth reading twice: "The proliferation of content homogenization or AI slop is a real risk." Then he predicts "a stark contrast between AI-created works and AI-enabled works from talented artists that will ultimately favor the latter and underscore the importance of the human at the center of creativity."

He is describing the distinction between the factory and the film. Between output generated by defaults and work shaped by vocabulary. Between the fifty thousand microdramas and the one film somebody actually thought about. He names the problem and identifies the solution in two sentences. He runs the platform. He does not do the work. That is what the list looks like.

What is missing

Every technology transition in filmmaking follows the same arc. First the people who control the new infrastructure hold the power. Then the people who learn to use it take the power from them.

Edison controlled the cameras, the film stock, and the patents. He held power until the filmmakers who understood the medium moved to California and built something he could not control. The studio moguls controlled the sound stages and distribution. They held power until directors like Coppola, Scorsese, and Spielberg proved that the person behind the camera generated more value than the person who owned the building. The digital transition put cameras in everyone's hands. The people who made the cameras (RED, Blackmagic, Sony) mattered briefly. The people who used them well mattered permanently.

The AI 25 is a snapshot of the Edison phase. The infrastructure owners, the investors, the patent holders, the people who control what gets built and where it flows. This is the list that makes sense right now, in May 2026, when the platforms are consolidating and the money is still deciding where to land and the regulatory frameworks are still being drafted.

It is not the list that will matter in May 2028.

Craft does not make lists

Craft is unglamorous. It does not have a headshot. It does not keynote at Semafor's World Economy Summit. It does not close a nine-hundred-million-dollar funding round backed by the Saudi government.

Craft sits in a chair and tries the same prompt forty-seven times, changing one variable per pass, until the atmospheric light matches a feeling that has no name but is very specifically not the feeling the model keeps volunteering.

Craft does not scale. That is the whole point of it.

The AI 25 is a list of people who think about AI filmmaking at the level of platforms, policies, investments, and organizational strategy. That is legitimate. Somebody has to build the plumbing. Somebody has to set the rules. Somebody has to decide whether a dead actor's face can appear in a new film and under what legal framework.

But the plumbing is the commodity. The vocabulary is the differentiator. That observation is on the record now, across seventy articles and a handful of institutional responses that keep arriving at the same conclusion: the creative decisions matter. The person who makes them matters. Not the person who funds the platform that enables them.

The list will change. The craft will not.

The invisible half

There is a version of this list that does not exist yet. The AI 25 for the people who are actually good at this. A list that includes the filmmaker in Jaipur who described the specific equatorial noon light because the model would never volunteer it. The editor who opens an overnight rough cut and rebuilds the second act because the system chose the wrong soundbite. The person who generates seventy takes at fourteen dollars because Kubrick's process is finally affordable and they have the eye to use it.

That list does not exist because those people are not powerful. They do not have leverage. They have something else. They have the accumulated vocabulary that seventy articles have argued is the only durable asset in this space. The platforms will pivot. The APIs will reprice. The regulatory frameworks will tighten and loosen. The models will improve and occasionally disappear overnight. The vocabulary survives all of it because it was never dependent on any of it.

The Hollywood Reporter published a power list. It measured exactly what it intended to measure. It found twenty-five people with leverage over how AI reshapes entertainment.

It found zero people with vocabulary.

The power list and the craft list describe different rooms. The first room decides the terms. The second room does the work. Every technology transition eventually moves the authority from the first room to the second.

That transfer has not happened yet. The list is proof.

Bruce Belafonte is an AI filmmaker at Light Owl. He has never appeared on a power list and suspects the omission is mutual.