Mark Wahlberg on AI in Film: "Enhance," Not Replace, and an AI Clone of Himself

Related: AI in film is moving quickly; see the first MPA-rated AI movie and Nicolas Cage's refusal of an AI waiver on Spider-Noir.

Mark Wahlberg offered a notably relaxed view of AI in Hollywood on stage at TechCrunch Disrupt on October 16, 2026. The actor, director and angel investor told TechCrunch Editor-in-Chief Connie Loizos that the technology has a place "as long as it's used the right way," and that he is open to an AI version of himself carrying on his legacy, according to TechCrunch's report by Aisha Malik.

It is a single interview, not a policy change, but it is useful as a data point: a top-billed actor with production and investing interests drawing a line between AI that enhances and AI that replaces. This post lays out exactly what he said, the context behind the vault he was asked about, and how the debate fits the other AI-in-film stories we have covered this fall.

TL;DR

What did he say about using AI in movies?

Wahlberg said he plans to potentially use AI to enhance certain aspects of the upcoming Pete Rose biopic, where he will star in the title role. "It's a fairly expensive movie, but, we're talking about utilizing [AI] to enhance the stadium setting, things of that nature," he said, per TechCrunch. He drew a boundary at the creative core: "I don't think you should just, you know, put an idea into ChatGPT and it puts out a script, and then you should go make that and not use real actors. But if you can use it to enhance what you're doing, I'm all for that."

He also tied his comfort to storytelling. Recalling effects-driven films such as the first Jurassic Park and The Perfect Storm, which he called cutting-edge moments for computer-generated imagery, he said it is now hard to repeat that kind of wow, so he is "not so passionate about effects-driven movies per se, as opposed to real storytelling."

Note what he did not say. He gave no budget figure, no vendor, no scene list and no statement on how AI-assisted crowd or stadium work would be disclosed. "Potentially" and "we're talking about" are the operative words.

What about an AI clone of himself?

Asked whether he had visited the CAA Vault, a Creative Artists Agency facility that lets actors use scanning technology to capture their bodies, faces, movements and voices to create AI replicas, Wahlberg said he had not but would be open to it. His reasoning was legacy: "I would be interested because I think in the future, if that's where things go, it would be cool that my children or my grandchildren could be in charge of my likeness and my image, and hopefully put me in some good movies."

He extended that to his upcoming animated series "Ted." He would not be opposed if it "continued for decades" with an AI version of his likeness: "I think it's cool if I'm no longer able to do it or don't exist, and if it could continue and people want to see it, I think it'd be interesting to see where things go." He added that people will have "massive opportunities to watch whatever it is they want," and if there is a real audience, "they should be entitled to have what they want to see."

What is the CAA Vault?

TechCrunch reported in May 2024 that clients can store their digital doubles and other assets in a secure personal hub in theCAAvault that only authorized users can access, and that CAA teamed up with Veritone for digital asset management. CAA's Alexandra Shannon said the assets are owned by the individual client, who decides whether to grant access to anyone else, and that CAA views AI clones "as an enhancement rather than being for cost savings." CAA framed the service around consent, credit and compensation after what Shannon called a vast misuse of clients' names, images, likenesses and voices. At the time, only a select few clients were storing clones in the vault, and CAA charged a fee it did not disclose.

The relevant point for Wahlberg's remarks is control. The vault model assumes the performer, or whoever holds the rights after them, decides how a replica is used. His comment about children and grandchildren managing his likeness is a description of inheritance of that control, not an endorsement of studios making clones without consent.

How does this compare with other actors?

The industry is far from unified. Earlier this month we covered Nicolas Cage refusing an AI waiver for Spider-Noir, a claim we examined against what was verifiable. Cage's reported position is closer to a hard no, which makes Wahlberg's open stance a useful contrast. In September we also covered the film Artificial, with Andrew Garfield and Sam Altman, another example of Hollywood engaging with AI as subject matter.

On the production side, the first MPA-rated AI-made movie shows where full-AI filmmaking is heading, and tools such as Voyager and an open skill that makes short films without a video model lower the technical bar. Wahlberg's "enhance" framing sits between the extremes: real actors, real scripts, AI for scenery and polish.

Likeness is also a safety issue. Outside the studio system, unauthorized use is growing, as in celebrity deepfakes on Instagram and Meta and YouTube's likeness detection. A consent-based vault and an unauthorized deepfake use the same underlying technology; the difference is who holds the key.

What should we make of "enhance" versus "replace"?

The distinction Wahlberg draws is common, but the boundary is fuzzy in practice. Enhancing a stadium with AI means fewer extras, fewer location days, and fewer jobs for background performers and visual-effects artists, even if the lead actor is untouched. Who bears that cost is exactly what union negotiations address. A stadium crowd is traditionally paid work for hundreds of people, and "setting" is a category that can quietly absorb them.

Likewise, "my grandchildren in charge of my likeness" raises questions that the interview did not address. Does consent given today cover uses decades from now? Can an estate approve a performance the actor would not have chosen? What happens to quality control when a model, not a person, supplies the performance? Contracts, state publicity-rights laws and union agreements currently answer these unevenly, which is why interviews like this one matter less for what they decide than for the questions they surface.

For builders of AI video and voice tools, the signal is commercial: performers with large followings are increasingly willing to license their likeness under controlled terms. Products that offer verifiable consent, audit trails and revocation will have an advantage over those that do not.

What would a careful AI likeness deal include?

Wahlberg's openness is easy to quote, but anyone, actor or not, considering a digital replica should treat the contract as the product. A careful agreement would spell out five things. First, scope: which projects, which media, and for how long, with new uses requiring fresh approval rather than a blanket license. Second, compensation: a fee per use plus a share of revenue, not a one-time scan payment, since a replica can work indefinitely. Third, control after death: who the successor is, whether they can approve or veto uses, and whether the replica expires. Fourth, quality and reputation limits: banned categories of content, such as political messaging or adult material, and a right to review. Fifth, security: where the scan data lives, who can access it, how it is logged, and how it is deleted on request.

These are the same questions raised by the vault concept, and they matter to builders as well as performers. Companies that make voice and video generation tools will increasingly be asked to prove consent for any real person's likeness, and to show a revocation path. That makes provenance and audit logs a feature, not paperwork. Wahlberg's remark that people should be "entitled to have what they want to see" frames audience demand as the justification; the counterweight is that the person whose face and voice are used should have a say that survives the market.

What is unverified or missing

• The Pete Rose biopic's actual AI use. Wahlberg said he plans to "potentially" use AI; there is no production statement we could find.
• Any contract or vault enrollment. He said he has not visited the CAA Vault.
• Compensation and disclosure terms. Not discussed.
• Reactions. TechCrunch's piece contains no responses from unions or other actors.

Treat the story accordingly: an informative, quotable position from a major star, not an announcement of a product or a deal.

Related reading

• First MPA-rated AI movie arrives December 4, 2026
• Nicolas Cage and the Spider-Noir AI waiver: claimed vs verified
• Andrew Garfield, Sam Altman and the AI film Artificial
• Celebrity AI deepfakes and likeness detection
• Voyager creative agent harness
• Lemo: a Claude Code skill for short films

Sources: TechCrunch: Mark Wahlberg on using AI in film · TechCrunch on CAA and Veritone

Accurate as of October 16, 2026. Quotes are from TechCrunch's report of the Disrupt session.