Hollywood has been abuzz with excitement and concern lately, with two prominent voices offering nearly diametrically opposed viewpoints.

One is Nicolas Cage. On October 10 at New York Comic Con, the ever-unpredictable actor took the microphone and revealed something he had been keeping under wraps: During the filming of Amazon Prime Video’s Shadow Spider-Man, the producers unexpectedly presented him with an “AI licensing waiver” in the midst of production.

Cage was blunt in his response: “I refused to sign that waiver. I wouldn’t permit them to utilize AI in Shadow Spider-Man.”

The other voice belongs to Mark Wahlberg. Just days after Cage’s “showdown” with Amazon, Wahlberg granted an exclusive interview to TechCrunch, adopting a markedly different stance: AI should be employed, but “in the right manner.” He even openly stated that he wouldn’t object to an AI clone continuing his legacy after he steps away from the spotlight.

One views AI as a monster to be kept at bay, refusing even to sign related documents; the other actively contemplates a “digital double” as his successor. Within the same industry and week, two contrasting approaches have surfaced.

Interestingly, these seemingly opposing stances converge on the same question: When AI can replicate a face, a voice, or an acting style, what remains for actors to hold onto? The answer lies not in technology itself but in human agency.

Is Wahlberg’s “right way” truly correct?

Let’s begin with Wahlberg. In his interview, he didn’t pretend to be deeply philosophical—his approach was pragmatic. He acknowledged AI’s profound integration into the film industry and argued that burying one’s head in the sand is futile; instead, the focus should be on how to utilize it. His “right way” can be distilled into a few principles: authorization, controllability, and serving creativity rather than replacing it. As for allowing an AI clone to carry on his legacy, he not only doesn’t resist it but is willing to take the initiative—provided the digital double is “personally handed over” by him.

Consider the underlying logic: Wahlberg isn’t fearless about AI; he fears it “growing uncontrollably”—becoming a weapon wielded by others without his consent. He’s willing to pass the baton because he wants to control its boundaries, pace, and even the direction of its next performance.

This fear is shared with Cage, though their solutions diverge. Cage chose to slam the door shut; Wahlberg opted to keep the key. Who is “right”? There’s no definitive answer. But one harsh reality Wahlberg sees more clearly than Cage: The door can’t remain shut indefinitely.

In the past ten days, AI has already surpassed the actors

What was happening in the outside world while Wahlberg spoke?

At dawn on October 18, actor Wu Yue angrily called out an account named “Hua Jie Beauty Diary” on social media for using an AI-generated “Wu Yue” to appear and voice a skincare product ad, labeled as “endorsed by Wu Yue.” The same account featured a video claiming “endorsed by table tennis champion Deng Yaping.” Deng’s team responded: no endorsement, no authorization, evidence being collected. The video even boldly displayed a disclaimer: “Author’s note: Contains AI-generated content”—as if that justified profiting off someone else’s likeness. The account was eventually banned, but Wu Yue’s “outrageous” remark sent shivers through the industry.

This is AI’s most menacing aspect: It doesn’t require your approval—or even your presence. A few minutes of public footage, a few lines of code, and “you” are selling products for others, with no power to stop it.

Zoom out further. In late September, Bona announced that China’s first AI hyper-realistic theatrical film, Sanxingdui: Tales of the Future, would be released nationwide on October 23, with a duration of 100 minutes and a National Film Bureau screening permit. In China, actor Chen Yilong signed a contract in August licensing his facial images to a studio—he sat in front of a camera making expressions while AI generated a “digital double” to act in AI films. His candid remark: “If you can’t beat it, join it.”

In Hollywood, Netflix co-CEO Ted Sarandos revealed in October that generative AI workflows have been integrated into roughly 300 Netflix productions, covering concept design, pre-visualization, post-production, and delivery—most intensively in post-production, where AI quietly handles tasks like crowd expansion and historical battle scenes that were once costly and time-consuming. This coincides with Netflix’s massive layoffs. Earlier, reports noted that as AI-generated vertical short dramas flood the U.S. market, crew recruitment for micro-short drama professionals in Los Angeles has plummeted by 80%, with some production companies jumping straight from live-action to “full AI.”

Notice something? While Cage refused to sign at Comic Con and Wahlberg discussed the “right way,” AI had already bypassed them both—and all debates about “whether to engage”—to storm into film sets, livestreams, and editing rooms. The window for argument is closing at a visible pace.

The coders are “legislating” for actors

Buried in Wahlberg’s “right way” is the industry’s greatest fear: Who defines “right”?

Two groups are racing to provide an answer. One is capital and platforms, driven by scripts of “faster, cheaper, more productive”—Netflix’s 300 productions are proof. The other is law and regulation: Last month, China’s Supreme People’s Court issued Guidelines on Lawfully Adjudicating AI-Related Disputes, clarifying that using AI to process a natural person’s likeness or generate a recognizable virtual image without consent constitutes infringement of personality rights like the right to likeness. In other words, the boundaries of “AI actors” are being defined by court rulings and judicial interpretations, not by actors themselves.

This is the truly unsettling part of Wahlberg’s interview. He might be one of the few Hollywood actors who understands this: As AI begins replicating “humans,” an actor’s labor is shifting from “selling time and body” to “managing and licensing a set of increasingly valuable digital assets.” Whether you sign or not, those assets exist—priced by capital, scheduled by platforms, measured by law. The difference is whether you hand over the baton or have it seized.

Cage’s “no” is about dignity; Wahlberg’s “I’ll do it myself” is about clarity. Dignity deserves respect, but clarity ensures survival.

A final observation

Writing this, I recall James Cameron’s comment last year: No generative AI was used in producing the new Avatar film. The man who pushed film tech to its limits guarded tradition with the most rudimentary methods. Meanwhile, the world’s strongest computing power is being used to generate hyper-realistic “battle scenes.”

AI’s entry into film isn’t a “yes or no” choice but a “who, how, and who owns it” allocation problem. Wahlberg at least got the first half right: Handing over the baton voluntarily beats having it taken. As for the second half—defining the boundaries, copyright, and legacy of digital doubles—he can’t answer it alone, nor can Cage.

This requires the entire industry to sit down and negotiate, page by page, the legal weight of “authorization” and the industry standards for the “right way.” The slower they negotiate, the faster AI will outpace them. By the time “I just” becomes “already,” there might not even be a seat at the table.

Wahlberg chose to sit at the table. That might be the “rightest” thing he’s done as an actor.