Raul Delapena Setiawan
The creator economy, a multi-billion-dollar industry built on the foundation of authenticity, personality, and human connection, is currently facing an existential threat. As generative artificial intelligence (AI) evolves from a curiosity into a highly capable commercial tool, a new, volatile fault line has emerged. On one side of this divide stand the "authorized" clones—digital twins licensed by creators to streamline brand deals, interact with global fanbases, and attend meetings in their stead. On the other side lie the "wild" clones: unauthorized, AI-generated facsimiles trained on the personal data and content of creators without their consent, knowledge, or compensation.
This technological arms race is not merely a legal hurdle; it is a fundamental redefinition of what it means to own one’s identity in the digital age.
The Case of Tana Mongeau and the Rise of "Wild" AI
The vulnerability of even the most established digital stars was laid bare on June 4, when creator Tana Mongeau—a titan of the industry with a combined reach of nearly 20 million followers across Instagram, YouTube, and TikTok—took to X (formerly Twitter) to voice her alarm. Mongeau shared a promotional video from Miso Labs, an AI voice generation company that touted its "Miso One" model as "the most emotive voice model in the world."
In the video, Mongeau observed an AI character that appeared to be modeled directly after her cadence, tone, and vocal mannerisms. "Why is this like my voice? Help," she wrote. Her distress resonated across the industry, highlighting the ease with which bad actors can now strip-mine a creator’s years of content to synthesize a digital puppet.
The incident was compounded by Aoden Teo, a co-founder of Miso Labs, who had previously boasted on social media that the company could "clone any voice with just 10 seconds of audio." In the same thread, Teo confirmed the existence of a model based on creator Salman Khan, flippantly adding, "Pls don’t sue us." Despite the public outcry and requests for comment from both Mongeau and Miso Labs, the video remained live, underscoring the lack of immediate recourse for creators whose likenesses are commandeered by emerging tech startups.
A Chronology of the Synthetic Shift
The rapid acceleration of AI likeness generation has moved from the realm of science fiction to a daily reality for content creators.
- Early 2024: The industry sees a surge in "official" digital twins. Influencers begin experimenting with AI avatars for multilingual content, allowing them to scale their reach without the logistical nightmare of manual dubbing.
- May 2025: Napster pivots toward an "agentic" AI video chat service, signaling a shift toward enterprise-level, consent-based digital twins meant for professional productivity rather than content creation.
- June 2025: The "Tana Mongeau incident" brings the issue of unauthorized cloning to the mainstream, forcing a conversation about the difference between "training" an AI and "stealing" a persona.
- Late 2025/Early 2026: A wave of high-profile creators, including Taylor Swift and Matthew McConaughey, begin filing aggressive trademark applications for specific vocal phrases, catchphrases, and visual clips in an attempt to create a legal "moat" around their identities.
Supporting Data: The Mechanics of Mimicry
The technology driving this shift is moving at a pace that far outstrips existing intellectual property law. Large Language Models (LLMs) and advanced audio-visual generation tools are no longer just processing text; they are "ingesting" the entire internet.
Gautam Goswami, CEO and chief AI officer at POP.STORE and ECHO-ME.AI, argues that the problem is systemic rather than incidental. "Creators need to understand that the biggest copier of them is Google Gemini," Goswami notes. "Nano Banana [powered by Gemini 3] has literally sucked up the pictures of every public person on the internet and created a model based on that."
The technical reality is chilling: the models are now so efficient that they require a "light" input—sometimes as little as 10 seconds of audio or a handful of high-resolution frames—to create a persistent, convincing digital twin. This means that a creator’s entire back catalog on YouTube or TikTok acts as a buffet for AI developers, who can scrape years of content to create a synthetic version that arguably knows the creator’s "voice" better than a casual viewer.
Official Responses and the "Ownership" Debate
The industry is currently split into two camps: the "open" advocates and the "protectionist" camp.
The Proponents of Managed AI
Some companies are positioning themselves as the ethical alternative to the "wild" AI landscape. POP.STORE and Napster emphasize a "closed-loop" ecosystem. According to Goswami, the profiles built on the POP.STORE platform are not shared outside their ecosystem and are designed to be completely deletable. "When they delete their account, it’s gone, gone," he says. "It’s not like if they go and train their face on Meta’s AI. If they do, there’s no way of taking it back."
Sean Morrison, a communications manager for Napster, echoes this sentiment. The company’s focus is on "consequences," ensuring that the rights of the individual remain front and center. "The industry is in a moment where it’s full speed ahead," Morrison says. "We take a different tack where those consequences have to be front and center."
The Skeptics and the Failed Deals
Not all attempts to monetize digital twins have been successful. The high-profile $975 million deal between TikTok star Khaby Lame and Rich Sparkle Holdings serves as a cautionary tale. While the deal promised a future of AI-powered brand integration, the subsequent 90% collapse in the company’s share price and the total silence from Lame suggest that the market for AI avatars is, at best, unproven and, at worst, a massive risk to a creator’s brand equity.
Implications: The Legal and Strategic Minefield
The legal landscape remains dangerously murky. Frank Poe, a lawyer specializing in creator rights, notes that current trademark law was never intended to handle the nuances of AI identity.
"It’s very difficult to prove," Poe explains. "You have to do discovery, you have to attempt to subpoena, to uncork how this model was trained and where they got sources from." For the average creator, the cost of litigating a copyright infringement claim against a well-funded AI startup is often prohibitive. While creators like Matthew McConaughey are attempting to "trademark themselves," this is an expensive, intensive process that is not a viable solution for the vast majority of independent influencers.
The Brand Perspective
For brands, the situation is equally precarious. While AI clones offer the allure of 24/7 engagement and lower production costs, they introduce a significant "reputation risk." An AI clone, if not constrained by strict guardrails, can hallucinate or go off-script, potentially damaging the reputation of both the creator and the brand partner.
Aleksei Poliakov, a global influencer marketing strategist, suggests that for the near future, AI clones will remain a "specialist tool." The consensus among experts is that until there is a clear, international legal framework—similar to the legislative efforts in Denmark to codify personal image rights—brands should approach AI partnerships with extreme caution.
The Road Ahead: Regulation or Chaos?
As the technology continues to democratize, the divide between those who can afford to protect their image and those who cannot will likely widen. Without governmental intervention, the "wild west" of generative AI will continue to siphon attention, revenue, and agency away from the creators who build the platforms that these very AIs are trained on.
The next two years will be defined by the "no training" clause. Expect to see creators, agents, and managers pushing for explicit language in brand contracts that forbids the use of their likeness to train generative models. For now, however, the digital doppelgänger remains a volatile, largely unregulated force—one that requires creators to be as vigilant about their data as they are about their content.
Ammar Sabilarrohman
Muslim
