EP7: Protecting Talent in the Age of AI: Rights, Likeness & GenAI with Ben Woollams, TrueRights

A New Category of Rights: Protecting Talent's Digital Presence

Drawing on eight years in influencer marketing, Ben built TrueRights to address a gap he saw forming across the creator economy — from micro- and nano-influencers to major global celebrities. Historically, brand-talent relationships were fairly simple: a creator would produce one piece of content and license limited usage rights. Today, brands want to repurpose content across websites, programmatic advertising, and paid media — and content licensing has struggled to keep pace, with little tracking, governance, or standardised pricing to reassure talent about how their content is being used.

Generative AI has intensified this problem. As content becomes dramatically easier and cheaper to produce at scale, brands are increasingly interested in using a person's "likeness" — generating entirely new AI-created images or videos of real talent — often described in the industry as building a "digital twin."

How Good Are AI Tools at Replicating Real People?

Ben describes the current state of AI-generated likeness as remarkably advanced "in the right hands." He shares an early example from working with TikTok's AI content tool, TikTok Symphony, which could generate branded video content but initially struggled with safely and accurately representing real talent.

TrueRights now partners directly with talent to licence approved likeness data, producing AI-generated imagery so accurate that talent themselves are often stunned by the results — right down to details like facial scars. However, Ben is clear that accuracy depends heavily on data quality: much like prompting a language model, feeding these tools inconsistent or incomplete reference material can lead to distorted or "hallucinated" results. Image and audio replication are relatively advanced, while accurately capturing distinctive mannerisms — the way someone walks, talks, or carries themselves — remains a harder technical challenge for video.

From "Intellectual Property" to "Intellectual Personality"

One of the most striking ideas Ben shares is a reframing of what's actually being licensed. Rather than simple intellectual property, GenAI likeness tools increasingly require deeply personal data — height, weight, mannerisms, accent, and more — leading some in the industry to describe this as "intellectual personality." This distinction underscores why licensing a person's likeness requires a fundamentally different, more careful approach than licensing a creative asset like a song or image someone has made.

Building a Licensing Framework for the AI Era

True Rights has built partnerships with nine different generative AI tools and production platforms, mapping the specific data requirements each needs to generate compliant, high-quality likeness content. Central to this is the company's proprietary "TrueRights stamp," which works alongside the C2PA content credentials standard to embed licensing information directly into content — allowing third parties to see exactly what usage rights and parameters apply.

The framework rests on two core principles: content must be fully copyright-approved, and every use must be authorised directly by the talent as rights holder. Looking ahead, Ben describes early plans for deeper integrations directly within generative AI platforms themselves — allowing brands to access properly licensed talent likeness at the point of content creation.

Brands Are Starting to Engage Responsibly

Ben points to encouraging signs that brands are approaching AI likeness responsibly, citing a recent conversation with H&M around their "digital twin" initiative — an approach designed not just to protect and remunerate the models involved, but to consider the wider production crew, including how they might contribute to reviewing AI-generated content rather than being displaced entirely.

Why Licensing Talent Likeness Differs from Music and Art Licensing

Unlike music, which has an established (if imperfect) licensing and royalties infrastructure built over decades, AI likeness rights represent largely uncharted territory — giving companies like TrueRights an opportunity to help shape industry standards from the ground up, rather than attempting to unpick decades of existing practice. Ben argues that likeness rights carry a distinct level of personal sensitivity compared to other creative IP, since brands aren't simply using someone's output — they're recreating the person themselves.

Global Perspectives: The US, Europe, and the UK

Ben highlights a notable difference in appetite for AI licensing solutions across regions. The US, he suggests, has adopted a more ambitious, fast-moving approach — partly driven by the entertainment industry's early exposure to these issues through events like the SAG-AFTRA strikes. France is emerging as a thought leader in Europe, while Ben describes the UK as comparatively cautious, with more hesitancy around decision-making, despite strong underlying interest and a track record of producing innovative companies in the space, including UK-founded generative AI platform Synthesia.