AI ‘content creators’ are getting harder to spot

AI-generated social-media personas are becoming harder to distinguish from human accounts as the technology improves and platforms rely on labeling policies that leave synthetic influencers in a regulatory gray area, according to a new analysis [1]. Early virtual influencers such as Lil Miquela, Imma, and Shudu Gram were visibly digital productions that required studios, money, and coordination [1]. Over time, characters like Emily Pellegrini and Aitana Lopez moved closer to the polished aesthetic of professional human influencers, posting from restaurants, festivals, and sporting events [1]. Pellegrini’s creator, who uses the pseudonym Professor EP, previously managed OnlyFans accounts and now sells courses on building AI influencers [1]. Lopez is managed by the Spanish agency The Clueless, which operates a roster of synthetic creators [1]. The tools for producing these personas have become widely accessible. Mainstream products from Google and OpenAI sit alongside specialized services from firms such as Higgsfield, HeyGen, and ElevenLabs, allowing almost anyone to create an AI influencer without a studio or significant expense [1]. The underlying technology draws on generative AI systems that, since the 2020s, have been capable of producing images, audio, and video from text prompts [3]. Google’s Gemini model, for example, is trained to process and generate text, images, audio, and video simultaneously [4]. Platform policies have not kept pace. Most major services now maintain rules covering synthetic media, but beyond requiring labels for AI-generated content, the policies largely slot the material into existing categories such as scams, spam, or impersonation [1]. AI personas that disclose their synthetic nature and do not impersonate a real individual often fall outside those categories [1]. YouTube, TikTok, and Instagram have developed labeling rules while simultaneously promoting their own AI tools, including some that can clone or simulate users [1]. TikTok, which surpassed two billion mobile downloads by April 2020, uses recommendation algorithms to connect creators with audiences, a system that can amplify both human and synthetic accounts [2]. Market estimates underscore the scale of the phenomenon. Some research firms project the virtual influencer market could be worth more than $60 billion by 2030, up from approximately $12 billion this year [1]. The ecosystem now includes AI influencer awards, beauty pageants, and dedicated talent agencies [1]. Below the high-profile avatars tracked by databases such as Virtual Humans lies what the analysis describes as an ocean of accounts flying under the radar [1]. Regulatory pressure is building incrementally. Europe’s AI Act will require deployers of generative AI systems to clearly disclose AI-generated or manipulated content, which could force platforms to strengthen flagging mechanisms or face fines [1]. Broader discussions of AI regulation have intensified as ethical concerns and harms from AI use have drawn attention from lawmakers [3]. In February 2025, Alphabet removed guidelines from its public AI ethics policy that had previously ruled out applying AI technology to uses likely to cause harm [11]. The analysis notes that the current regulatory focus remains on content rather than on whether the account posting it represents an actual person [1].