AI influencers are computer-generated characters who post, collaborate with brands, and build communities like human creators. On Instagram, they are no longer a gimmick: Lil Miquela has existed since 2016, Aitana López surpassed 300,000 followers according to Time in 2024, and Meta already labels some AI-generated content. The real marketing question is: should you work with them, and under what conditions?
AI influencers: what they are already changing on Instagram
An AI influencer is not just a pretty image produced by Midjourney or Stable Diffusion. It is a scripted character, with a visual identity, editorial voice, posting routines, and sometimes a full team behind each post. The difference from a classic brand avatar? It borrows the codes of influence: selfies, stories, partnerships, lifestyle, comments, short-form content.
Since 2015, I’ve watched Instagram move from a polished square feed to Reels, and then to TikTok-style algorithmic recommendations. In this context, AI influencers enjoy a clear advantage: they are available, controllable, multilingual, and have no shooting constraints. But they also have a weakness many advertisers underestimate: without human tension, without imperfection, engagement can quickly feel like a cold curiosity.
The historical example remains Lil Miquela, also known as Miquela Sousa. Launched on Instagram in 2016 and associated with the company Brud, she was named by Time in 2018 among the “25 Most Influential People on the Internet,” precisely because she was already blurring the line between avatar, fiction, and influence. Eight years later, that line has become a matter of compliance, branding, and trust.
Why brands are closely watching AI influencers
For an advertiser, the first appeal is control. No delays in a shoot, no bad buzz tied to an unexpected personal statement, no weather canceling a production. A brand can adapt the same character into an Instagram carousel, Reel, vertical TikTok video, e-commerce visual, or LinkedIn activation, with a consistency rarely achieved with human casting.
The second appeal, more subtle, concerns rights. A human influencer negotiates usage duration, territories, industry exclusivity, and paid media use. With a proprietary or co-developed character, the brand can secure more assets, provided the contracts cover the tools, prompts, models, synthetic voices, and image licenses. It’s technical. It’s also where profitability is decided.
In the campaigns we see, it’s better to use an AI influencer for worlds where art direction matters more than lived proof: fashion, conceptual beauty, gaming, entertainment, innovation, futuristic products. For a health, parenting, or personal finance recommendation, honestly, the virtual character needs to remain very tightly controlled. The absence of real experience quickly becomes a credibility issue.
The rise of synthetic content also forces social media teams to better understand platform signals. On Instagram, the algorithm still favors retention, quick interactions, and account consistency; if you work this channel, our analysis of mastering the Instagram algorithm usefully complements this discussion.
Real examples: Lil Miquela, Aitana López, Lu do Magalu
The well-known cases mainly show that there is not just one model of AI influencer. Some are independent narrative characters. Others are social commerce mascots, integrated into a brand. Still others become lab tests to measure public tolerance for artificial content.
| Character | Known origin | Strong platform | Public data to note | Marketing take |
|---|---|---|---|---|
| Lil Miquela | CGI avatar launched in 2016, associated with Brud | Cited by Time in 2018 among the 25 most influential people on the Internet | Foundational case: storytelling, fashion, pop culture | |
| Aitana López | AI influencer created in 2023 by the Barcelona agency The Clueless | More than 300,000 followers and up to €10,000 per month reported by Time in 2024, with an average around €3,000 | Proof of commercial interest, figures to be read as self-reported | |
| Lu do Magalu | Virtual character linked to Magazine Luiza in Brazil | Instagram and social commerce | Axios flagged it in 2026 as a virtual profile to watch; CheckBB indicated 8.9 million Instagram followers in 2026 | Brand avatar model, closer to retail media |
| Granny Spills | AI character featured in the media in 2026 | TikTok and Instagram | Time reported 400,000 TikTok followers and 1 million on Instagram in its first few weeks | Powerful novelty effect, durability still to be proven |
Be careful with the numbers. Followers are visible, but revenue, engagement rates, and campaign performance are rarely publicly audited. In 2026, several sources noted that AI influencer income often relies on statements from agencies or creators, not platform-certified data.
This is the classic trap: confusing media visibility with commercial effectiveness. An avatar may attract the curious for three weeks, then fade if its content lacks a community mechanism. The right metric is not “is the character impressive?” but “does it drive saves, shares, useful comments, and qualified clicks?”
Transparency, AI labels, and trust: the sensitive issue
Meta announced in 2024 that AI-generated images would be labeled on Facebook, Instagram, and Threads when detectable through standard industry indicators. Photorealistic images created with Meta AI must carry the mention “Imagined with AI,” along with visible markers, invisible watermarks, and metadata. This is not a cosmetic detail: it is a new layer of reputation.
In 2026, Instagram was also testing optional “AI Creator” labels for accounts that often create with AI, according to Engadget and Digital Camera World. Even if they remain optional, the debate does not go away. An advertiser that hides the synthetic nature of a profile takes a simple risk: being accused of manipulation once the community spots the setup.
My view is clear: for a paid collaboration, it is better to clearly signal that a character is virtual. Not necessarily in every sentence, but in the bio, partnership disclosures, and sensitive content. Audiences accept fiction when it is openly acknowledged; they punish deception when it resembles fake intimacy.
This requirement is part of a broader movement around social technologies, assistants, and AI objects. Brands exploring these uses will do well to also follow developments at Meta, for example around the Meta AI pendant and voice capture, because data and consent will weigh heavily on campaigns.
How to test an AI avatar without damaging your brand
A serious test starts small. No need to create a global synthetic spokesperson with 80 posts ready to go. The healthiest format is to launch a limited series: 6 to 10 pieces of content, a clear editorial promise, a precise product scope, and a comparison with a human creator over the same period.
The brief must be stricter than for a traditional influencer. You need to define the character’s story, speaking boundaries, off-limits topics, how to respond to comments, and the level of transparency. Without that, the creative team will produce a beautiful empty shell.
- Set the objective : awareness, traffic, lead generation, UGC-like content, or a creative territory test.
- Measure against a control : compare with a human creator of similar size or with your usual brand content.
- Check the rights : image, voice, prompts, model used, paid reuse, usage period.
- Display the AI nature : bio, partnership disclosure, clear wording in commercial content.
- Analyze the comments : curiosity, rejection, questions about the product, accusations of fakery, brand sentiment.
For human creators, the issue is not just a threat. The smartest ones will use AI to prototype concepts, create backdrops, translate formats, and test hooks, without giving up their face or lived experience. On YouTube too, tools are evolving quickly; the AI built into Ask Studio and connected TV use cases shows how platforms are looking to support creation without necessarily replacing creators, as explained in our piece on YouTube's new AI features.
Instagram, TikTok, YouTube: where do avatars perform best?
Instagram remains the natural territory for AI influencers, because controlled imagery still carries a lot of weight there: Reels, carousels, stories, brand collaborations, lifestyle aesthetics. A photorealistic avatar can establish itself there quickly if the art direction is strong. But the algorithm doesn't reward beauty alone; it takes rhythm, signs of conversation, and formats that make people want to come back.
TikTok is rougher. The platform loves weirdness, social commentary, gimmicky characters, and formats that trigger an immediate reaction. An AI avatar can blow up fast there, as shown by the Granny Spills case reported by Time in 2026, but interest fades quickly if the concept boils down to “look, this isn’t a real person.”
YouTube calls for a different depth. Shorts can serve as a launchpad, but a lasting channel requires storytelling, a voice, recurring appointments, and often a larger world. Twitch, for its part, can welcome avatars and VTubers, but live interaction quickly exposes setups that are too rigid. Conversational latency, repetitive responses, the character’s limits: everything shows.
LinkedIn deserves a separate treatment. An AI avatar can support monitoring or a brand identity, but it won’t replace the credibility of an executive taking a stand. For B2B strategies, it’s better to use AI as an editorial assistant rather than as a fake expert; our article on AI on LinkedIn in 2026 helps set that framework.
Micro-community trends also matter. Avatars work better when they speak to an identifiable micro-culture: digital fashion fans, anime, K-pop, gaming, futuristic beauty, social retail. To understand these niches and their codes, the topic of microcultures on social media is directly linked to the success of a virtual character.
ValueYourNetwork has been supporting brands, creators, and talent since 2015 in their social media strategies, influencer campaigns, and format choices. Whether you are an influencer or an advertiser, if you want to grow your social media with a solid, measurable approach, contact us.
FAQ about AI influencers
What is an AI influencer on Instagram?
An AI influencer is a character generated or animated by digital technologies, designed to post and interact like a creator. It can be entirely fictional, like a CGI avatar, or managed by an agency, a brand, or a creative team.
Do AI influencers actually make money?
Yes, some sign collaborations, but public earnings are often self-reported. Time reported in 2024 that Aitana López could earn up to €10,000 per month, with an average of around €3,000, while higher figures published afterward should be treated with caution.
Does Instagram require AI-generated content to be labeled?
In 2024, Meta announced labels for detectable AI images on Facebook, Instagram, and Threads, including “Imagined with AI” for photorealistic images created with Meta AI. In 2026, Instagram was also testing optional “AI Creator” labels.
Should a brand work with AI influencers?
Yes, if the goal is creative, experimental, or image-driven, and if transparency is clear. For topics based on personal experience, a human creator is often more credible.
Will AI influencers replace human creators?
No, not in the short term. They will mainly take a place in hybrid campaigns, while human creators will keep the edge in trust, lived experience, and community relationships.