Josslyn. 23.2M TikTok followers. She shares a ‘Sunday Reset’ weekly, models the newest TikTok trends, and supplies sponsored content for her favorite beauty, wellness, and snack brands. She goes to concerts, her favorite color is orange, and she loves to read. Her followers DM her questions, which she responds to enthusiastically, sharing links to products with a discount code.
She’s also not real.
AI-generated influencers have become a fraud scheme enabler, a passive-income scheme for anonymous users, an option for brand promotion, a catfishing tool, and a venue for political influence campaigns.
This new class of online actor creates real reputational risks, both for brands and for executives or staff who may get impersonated. Building a convincing AI persona now takes minimal time, money, and skill, meaning malicious actors are able to do this at scale. Video generation tools can produce increasingly authentic-looking “people”, and AI tools can scrape the internet for trending topics and popular creators to build profiles optimized to gain a following. This risk is further compounded by social media platforms’ limited regulations on AI-generated content, making these personas both difficult to detect when they first emerge and even more difficult to contain once they’ve gained a following.
An AI-generated influencer is a digital, computer-created persona managed by an individual, team, or software to post content, attract followers, and promote products on social media. These accounts often do not disclose that the persona is AI-generated, in part because some social media platforms and regulations do not require it.
Accounts like these began emerging as early as 2016, particularly as @lilmiquela—a verified, self-identified “robot” with 2.2M followers—gained popularity. However, the signs of digital generation were obvious at the time: cartoon-like figures and faces, pixelated backgrounds, poor voice-overs, etc. Today, the realism of image and video generation makes detection far harder for the average user or a time-crunched marketing team looking for its next influencer.
The barrier to entry has dropped as well, making it exceptionally easy for the average user to create a realistic, active persona across video platforms. Open-source models are readily accessible and social media users have increasingly shared their process for building AI-generated influencers.
Operators of synthetic accounts are often driven by opportunities to monetize. Reddit users have shared methods to generate passive income using AI-generated influencers, driven by follower amassment, paid chat engagement, affiliate marketing, and brand promotions. Some users also create cheap digital products (you guessed it, also generated by AI) and sell them on digital storefronts attached to their profile, or solicit donations using the popular “buy me a coffee” prompt in their bio or a Linktree.
Synthetic personas are also well positioned to perpetrate fraud schemes, particularly as deepfakes become increasingly realistic. Deepfake scams have escalated significantly in recent years; reported deepfake-fraud financial losses have reached at least $3.7 billion in 2026. For example, scammers have leveraged synthetic personas in the likeness of authentic influencers to promote fake brands or solicit donations, making it extra difficult for the average consumer to distinguish between trusted voices and malicious actors.
The difficulty in identifying synthetic persona operators opens the door for malicious actors, including state actors, to create profiles that push their ideology, criticize brands, or otherwise detract from authentic narratives across platforms. Personas with strong political, environmental, or social “beliefs” can be trained to optimize engagement within likeminded communities, making it easy to generate a following, increase credibility, and further entrench specific narratives that align with ideological interests.
AI-generated influencers are no longer a novelty, but an active liability. Brands using them face potential backlash for inauthenticity, and brands that aren’t actively monitoring for them face impersonation risks and potential reputational damage from personas operating in their name or industry. Some distinct risk vectors include:
Calls for AI Marketing Transparency: Alethea has observed growing pressure from consumers and authentic influencers for brands to disclose their use of AI-generated influencers for marketing. Some high-reach users have taken to exposing these AI-generated personas outright, advising consumers how to spot artificial personas, and highlighting products or brands being advertised by likely inauthentic accounts.
Brands that rely on influencer marketing risk eroding consumer trust if they intentionally (or unknowingly) utilize an inauthentic persona, particularly as scrutiny of these accounts grows.
The False Credibility Dilemma: Some personas claim expertise in certain fields. They purport to be doctors, nutritionists, trainers, lawyers, etc., and have obtained millions of followers, creating perceived credibility. Some accounts even appear verified, like Wellness.Jada (1.3M Instagram followers), who shares nutritional advice, cross-posts to “her” YouTube channel (43.7K subscribers), and advertises herbal remedies in her bio.
This, coupled with realistic videos and content, can easily and quickly convince users to trust a persona’s recommendations or warnings about products. Brands are particularly exposed when a persona posing as an industry expert criticizes their company or product, sometimes with financial incentive to promote a competitor instead.
Coordinated Activity: AI-generated accounts are cheap to create, fast to monetize, and easy to scale. Multiple personas can be spun up in unison to push the same message, whether or not that coordination is intentional. For example, in a recent investigation, Alethea identified at least 26 active Instagram accounts exhibiting coordinated behavior: advertising the same nutritional supplement (sometimes using the same link in bio token, suggesting the affiliate income tracks back to a single individual), using the same grammar and style, and intentionally varying persona archetypes to reach diverse audiences.
Impersonation Risks for Executives and Staff: These personas are often created using the likeness of real, existing users on the internet. In some cases, this could be the amalgamation of several users into one, including their faces, bodies, voices, and posting patterns. In other cases, it can be a near replica of a person’s identity and voice, impersonating them by name or with an alias. This creates an additional set of risks to executives and staff, who may unknowingly become the face of an alternate advertising campaign for competitors, unrelated or provocative content that sparks backlash, or even a critical voice targeting their own brand.
Niche Platform Exposure: While these AI-generated influencers have become most prevalent on social media sites hosting short-form video content like Instagram, Facebook, TikTok, or YouTube, the content is routinely cross-posted onto niche platforms, where it can circulate in echo chambers with less vetting and a higher propensity for extremist or violent content. Brands without visibility into these spaces risk missing how such communities are interpreting, interacting, and amplifying a persona’s messaging. These platforms are also slower to adopt AI-content disclosure requirements as regulation matures elsewhere, making the problem harder to see and police.