Extended Cut: Why XR Is Expanding Celebrity Payments to Cover AI Ad Performances

As AI-generated performances move rapidly from experimentation into mainstream advertising, the question is no longer if they’ll be used—but how they’ll be governed, compensated, and tracked.

With new provisions under the 2025 contract from SAG-AFTRA introducing formal categories for Digital Replicas and Synthetic Performers, the industry is facing a structural shift: AI talent now needs to function inside the same systems that protect human performers.

XR (Extreme Reach), known in the advertising industry for handling payments and rights for on-camera talent and production crews for advertising, is expanding its remit into one of the most contentious frontiers in media: AI-enabled performers.

The Measure recently sat down with Graham McKenna, CMO and Tim Hale, Managing Director, of XR to discuss the latest news that the company now supports payment of AI performers in advertising. You can read Jon Lafayette's piece covering the announcement last week here.

The following is an extended look at XR's thinking behind this expansion and the immediate needs to be responsive to AI's role in the market.

From traditional scale actors to A-list celebrities and professional athletes, XR already touches payments for more than 80% of U.S. ads featuring talent. Now, the same infrastructure that underpins commercial payments and rights management for human performers is being adapted to handle digital replicas and synthetic performers as they become more common in advertising.

“We’re known in the industry for paying actors, talent, celebrities,” said Graham McKenna of XR. “What we’re announcing is an expansion of that to cover what we’re calling AI-enabled talent that are going to be appearing in advertising more and more, beginning now and into the future.”

Defining AI-Enabled Talent: Digital Replicas vs. Synthetic Performers

A key part of the challenge is simply defining who or what counts as “talent” in an AI-driven world.

Extreme Reach distinguishes between:

• Digital replicas: AI-enabled performances based on the likeness or performance of a real, living actor or celebrity. These can also extend to deceased performers when their image or likeness is used with consent from an estate.
• Synthetic performers: Entirely computer-generated characters not based on any specific, known person. Think of fully fabricated celebrities that exist only in pixels and code.

Hale drew a clear line between the two:

“A synthetic performer would be entirely computer generated. An AI-enabled performer can be a live performer who’s been adjusted. You aren’t bringing them back for a separate shoot.”

The line can easily be crossed, he noted, if brands try to approximate a recognizable person too closely.

Why Now? SAG-AFTRA and a Market That’s Moving Fast

The timing of this expansion isn’t accidental. McKenna says the latest Screen Actors Guild–AFTRA (SAG-AFTRA) 2025 commercial contract was a major catalyst.

At the same time, the market is already inching toward AI-driven talent. McKenna shared early findings from an upcoming XR survey of marketers and found that around 22% are already experimenting with both digital replicas and synthetic performers. “We’re seeing a growing need in the market to be able to accommodate this.”

In other words, XR isn’t chasing a hypothetical trend; they’re trying to meet the market at the moment it begins to scale.

Provenance and Rights: The Hard Part Behind the Scenes

On some level, adapting the platform wasn’t an enormous conceptual leap. “On one level, it’s not overly complicated, because these are the issues that we manage for real performers,” said Hale. “A synthetic performer versus a real one — there’s a crossover for a lot of that.”

But one aspect is significantly more complex in the AI era: Provenance.

“There’s an element of provenance behind an AI-generated image or actor performance,” Hale explained. “Being able to know how it originated, whether or not you have the IP rights to that or if they’re shared with another owner, is going to be critical.”

Extreme Reach already tracks talent rights “from the production set all the way through ad delivery,” McKenna said. That includes whether the brand or agency has the right to air a spot, where it can air, for how long, and on which platforms.

Now that same rigor needs to apply to both digital replicas and synthetic performers, including future scenarios where a celebrity’s likeness is remixed endlessly into thousands of versions of an ad.

“You’re going to have versions of celebrity talent remixed into different versions of ads, and that needs to be tracked all the way through to play-out as well,” McKenna said.

How Payments Work for AI-Enabled Performers

One of the biggest questions for advertisers and performers is how payment and compensation will function in this new environment.

McKenna explained that when a digital replica is based on a living actor, it’s treated like the performer themselves in terms of compensation. For deceased talent, compensation is managed via the estate or rights holder.

For major celebrities, Hale underscored that this is squarely in the realm of big-ticket negotiation. “Celebrities are negotiated. Every celebrity, living or dead, iis a negotiated arrangement with whoever owns the rights to that celebrity.”

In other words, this is not a flat-fee, plug-and-play environment. Like traditional endorsements, it’s highly bespoke, driven by negotiation between the rights holder and the brand or agency, with Extreme Reach handling the translation into enforceable payment and rights structures.

Scaling AI for Global and Local Campaigns

One of the most compelling promises of AI-enabled talent is its potential to scale creative more efficiently, especially for global brands.

Hale pointed to the opportunity to localize existing creative without bringing talent back for new shoots:

A brand can run an English-language commercial and then “localize it in a number of different languages, because they’re a global company. They don’t have to reshoot with native speaking performers. They can leverage AI to localize it.”

That same logic applies within a single country. Hale cited the example of a major automotive brand with 1,500 dealerships:

They could “leverage AI to have that well-known actor — that voice of the brand or image of the brand — be localized for every one of those dealerships… It’s creating a huge opportunity from the creative side.”

But every one of those thousands of localized variations has to be tracked, both for payment and compliance. That’s where XR believes it has a defensible edge.

Union vs. Non-Union in an AI World

As AI expands, questions about union vs. non-union work are becoming sharper, especially as budgets tighten and digital content proliferates.

Hale explained that when an agency or brand signs on as a SAG-AFTRA signatory, they agree to use union performers for covered advertising content. But many brands today are not direct signatories; instead, their agencies or production partners are.

XR, for its part, is agnostic:

“We work with brands and agencies who work union or non-union,” Hale said. “The key element beyond paying the performers and the crew is really around rights management. And so regardless of whether it’s union talent or non-union, you still need to track all the rights.”

That need only intensifies in an AI-heavy environment where the line between “real” and “synthetic” is increasingly blurred.

Multiple Owners, Shared IP, and an Unsettled Legal Landscape

Hale likened future ownership of synthetic performers to songwriting, where multiple composers and publishers share rights to a single work.

Referring to synthetic personalities like Tilly Norwood, he noted they “were actually developed by a number of people, so they would each have their own interest in the copyright for that creation.”

That model implies that synthetic performers may have complex, fractional ownership structures, potentially involving creators, modelers, animators, and AI engineers — each with a stake in the underlying IP.

What’s clear, he suggested, is that awareness of digital likeness rights is rising fast and legal standards are still evolving.

A Rights-First Approach to an AI-Heavy Future

As the volume of creative explodes from tens of versions of an ad today to hundreds or thousands tomorrow, the risk of confusion, misuse, and rights violations grows in lockstep.

“There might be 50 to 100 versions of an ad today,” McKenna said. “And to be able to track what’s real and what’s not becomes really, really important.”

In that environment, Extreme Reach is positioning itself not just as a payments processor, but as a rights and provenance backbone for the next era of advertising — one where the “talent” on screen may be a person, a replica, or something entirely synthetic, but the financial and contractual obligations are every bit as real.