Your Fave Influencer Might Be a Ghost in a Machine — and Somehow That's Less Creepy Than the Alternative
Let's set the scene. You've been following this creator for about three months. Their content is consistent, their aesthetic is immaculate, their opinions feel surprisingly nuanced for someone who also posts thirst traps. They respond to comments. They have a personality. You've started to think of them as, like, a person.
And then someone in the comments drops a link and your whole afternoon falls apart: this creator is not a person. This creator is a language model wearing a face that was generated by a different model, and the "authentic behind-the-scenes" content you've been watching is the output of a prompt someone typed into a chat interface at 2 a.m.
Here's the question that should be keeping you up at night, though: does that actually make them worse than the alternative?
Welcome to the Uncanny Valley, Population: Your For You Page
The uncanny valley is that uncomfortable zone where something looks almost human but not quite — close enough to trigger recognition, wrong enough to trigger revulsion. Early CGI faces lived there. Certain wax museum figures live there. And now, increasingly, AI influencers are setting up permanent residence there, except the valley keeps getting shallower.
The technology has gotten genuinely good. We're not talking about the obviously-fake CGI models from five years ago that looked like they were rendered on a PlayStation 3. We're talking about generated personas that move, speak, react, and post with enough specificity that the average scroll session won't catch the seam. Some of them have origin stories. Some of them have "opinions" on current events. Some of them have, and this is where it gets truly unhinged, fan communities that know they're fake and don't care.
Lil Miquela has been doing this since 2016. She has millions of Instagram followers, brand deals with Prada and Calvin Klein, and a fanbase that has collectively decided her synthetic nature is a feature, not a bug. She's not the anomaly anymore. She's the blueprint.
The Honesty Paradox Nobody Wants to Acknowledge
Here's the thing about human influencers that we've all tacitly agreed to forget: they lie constantly, and we know it, and we watch anyway.
They lie about which products they actually use. They lie about their skin being naturally that clear. They lie about how long it took them to look that effortless. They perform mental health vulnerability in ways that are carefully calibrated for engagement. They manufacture conflict for content. They apologize for things they're not sorry about and express excitement for things they find deeply tedious.
This is not a callout. This is just the job description. The entire influencer economy is built on a specific kind of managed performance that calls itself authenticity while being its precise opposite.
Now consider the AI creator. An AI creator cannot lie about using a product they don't believe in because — and stay with me here — they don't believe in anything. They have no off-camera life to misrepresent. They have no secret opinions they're hiding from their brand partners. They are exactly as much of a construction as they appear to be, which means there's a weird, backwards sense in which they're more honest than the human who's pretending not to be a construction.
The fake thing isn't pretending to be real. The real thing is pretending to be unmediated. Which one is actually deceiving you?
What Audiences Know and Choose Anyway
The really disorienting part of this whole situation isn't the technology. It's the audience behavior.
Studies — and there have been several — show that people often continue engaging with AI influencers after learning they're artificial. Not despite knowing. Not accidentally. After being told, explicitly, that there's no human on the other end, a significant portion of audiences shrug and keep watching. Some of them report feeling more comfortable after the reveal, because the parasocial pressure is gone. You're not worried about a bot's mental health. You're not projecting human need onto a language model. You can just enjoy the content without the guilt.
This is not irrational. This is a very reasonable adaptation to an environment where parasocial relationships with human creators have become genuinely exhausting. The AI influencer is, in a very real sense, the low-maintenance option.
The Brand Deal Problem (And It's a Big One)
Okay, but here's where the synthetic creator thing gets genuinely thorny, and we should sit with the discomfort instead of optimizing it away.
If an AI influencer promotes a product, who is responsible for that endorsement? There's no human experience behind it. The "creator" didn't try the supplement or test the skincare or actually feel the difference in their sleep quality after using the weighted blanket. The FTC disclosure rules that apply to human influencers were built around the concept of a person having an actual relationship with a product. When the person is a prompt and the relationship is a marketing brief, the whole framework collapses.
This is where the "more honest" argument hits its ceiling. Sure, the AI isn't lying about its personal life. But it's potentially being used to simulate endorsement in ways that are structurally deceptive in a completely different direction. The synthetic personality becomes a laundry machine for advertising. Clean, frictionless, legally ambiguous.
So What Do We Actually Do With This?
The emergence of AI creators isn't going to stop because we find it philosophically uncomfortable. The economics are too good, the technology is too accessible, and the audience appetite — bewildering as it is — is clearly there.
What it probably should do is force a long-overdue conversation about what we actually want from the people (and non-people) we follow online. If the answer is "consistent content that fits my aesthetic preferences and doesn't make me feel bad about myself," then AI creators are already delivering that. If the answer is "genuine human connection and the messy, unpredictable experience of following someone who is actually living a life," then no amount of generative sophistication is going to fill that gap.
The audience gets to decide. Which is both the most democratic thing about this situation and the most terrifying.
Because if we're being honest — and apparently that's what the robots are for now — a lot of us already made our choice. We just haven't admitted it yet.
