A "Person" You Can't Tell From Real: The Numbers Behind the Virtual Creator Economy

A few weeks ago I caught myself scrolling past a fashion post. Beautiful face, clean styling, a Calvin Klein tag in the corner. I almost kept going. Then I noticed the name in the bio: Lil Miquela. Not a real human. A rendered one.

And she's not some curiosity anymore. She's a line item in a brand budget.

I went looking for the numbers, because the hype around AI influencers has been loud and the receipts have been thin. What I found genuinely surprised me. So let me walk you through it.

First, the size of this thing

The virtual influencer market sits at $11.74 billion as of 2026. By 2032 it's forecast to hit $154.6 billion. That's a 41.29% compound growth rate, if you want the precise figure.

Let that number settle for a second. A hundred and fifty billion. For personalities that didn't exist as a category a decade ago.

If you cast the net wider and count "AI influencer" in the broad sense, you get $13.4 billion today heading toward $62.67 billion by 2030. The creator-tool layer alone, the software that lets one person run what used to need a studio, is $4.71 billion. The whole creator economy around it: $323.48 billion, on its way to $820.83 billion.

Why does this matter? Because money at this scale doesn't sit in the "experiment" bucket anymore. It's an industry. And industries attract serious operators, serious fraud, and serious regulators. Hold that thought.

What AI actually did to creators

Here's a misconception I want to kill early: AI is replacing creators.

It isn't. It's doing something more interesting. It's acting as what I'd call an operations layer.

Think about what a single creator used to need. A producer. An editor. A community manager. An analyst to figure out what to post next. That's a company. Now one person can carry all of it on a laptop, and the quality bar hasn't dropped, it's gone up.

The numbers back this up. As of 2026, 75% of professional creators use AI tools for planning, scripting, and editing. They produce roughly 40% more content than they did two years ago. Forty-six percent lean on AI for audience analysis. Thirty-four percent let it handle the admin drudge: community, email, moderation. That buys back about 15% of their time for the work that actually matters.

And here's the line that jumped out at me. Chief Marketing Officers are now steering up to 30% of their influencer budgets into virtual creators. Not humans. Virtual ones.

China is moving harder than anyone. They've put $1.6 billion into the virtual influencer space and they're running 340 million active followers through it. That's a U.S.-sized population, just watching rendered people.

The gap AI didn't close

Now here's where it gets uncomfortable.

Women make up the majority of creators on Instagram and TikTok. You'd think a tool that democratizes production would flatten the earnings gap. It hasn't.

Per NeoReach's 2025 report, women creators earn about $0.77 for every $1 men earn on comparable sponsorships. And the AI adoption curve is making it worse before it gets better: 68% of women creators use AI tools regularly, versus 79% of men. An eleven-point gap. It compounds.

The most telling stat, though, is who owns the upside of this boom. Among the top 25 virtual influencers by followers, roughly 80% are owned or operated by male-led studios. So the money flowing into "virtual women" is largely flowing to men who built them. Sit with that for a moment.

When asked why they're slow on AI, 31% of women creators cited "lack of time to learn new tools." Nineteen percent of men said the same. The barrier isn't interest. It's hours in the day.

The names actually making money

Let's get concrete. Who's cashing the cheques?

Lu do Magalu is still the benchmark. She pulled about $2.5 million in 2024 across 74 brand collaborations. That's roughly $34,320 a post, about forty times what a comparable human influencer charges. Forty times. Let me say that again, because I had to read it twice: forty times.

Behind her there's a real bench now. Lil Miquela has stacked an estimated $11 million in career brand-deal revenue working with Calvin Klein, Prada, Samsung. Imma pulls low six figures a month with IKEA, Coach, Calvin Klein. Noonoouri is signed to Warner Music and fronts Dior, Versace, Valentino. Shudu, the diamond-skinned model, goes for mid five figures a single post with Balmain and Fenty Beauty. Aitana López commands five figures monthly for a brand called Big.

What I want you to notice about this list isn't the dollar amounts. It's the brand names. These aren't trial-budget partnerships. Prada, Dior, Samsung, IKEA. That's core-marketing-spend territory.

The engagement question

So why are brands paying a forty-times premium?

Because virtual influencers, on average, outperform humans on the only metric that justifies the spend: engagement.

HypeAuditor puts the average virtual influencer campaign engagement at 5.67%, versus 1.89% for human creators. Roughly three times higher. Prada's Lil Miquela collaboration generated 30% higher engagement than the brand's campaign average. Thirty. Percent.

But, and this is the part most summaries skip, in categories where authenticity carries the sale, human creators still win. By up to 2.7×. So if you're selling skincare or a story about your own life, a real person still moves more units. If you're selling fashion, fantasy, or future-facing tech, the virtual creator pulls harder.

There's no universal winner. There's a category call to make.

How saturated is this, really

The honest answer: saturated at the top, wide open in the middle.

The number of virtual influencers with over a million followers jumped from about 150 in 2023 to more than 400 in 2026. Engagement on those mega-accounts has slipped from 5.9% to 5.67%. Small slide, but directional. The top is getting crowded and the returns are thinning.

But go down a tier and the picture flips. Virtual micro-influencers, the 10K to 100K follower range, are the fastest-growing cohort in the entire space. Brand adoption across surveyed companies climbed from 60% to 73% in two years. And in financial services? Only 22% of brands have run a virtual-creator campaign. Twenty-two. That's an open vertical, not a closed one.

By 2030, CMO budget share for virtual creators is forecast to hit 40 to 45%. We're nowhere near the ceiling.

The part brands won't say out loud

Here's something the data revealed that I keep coming back to.

Brands are happy to let AI find creators. They are not happy to let AI verify them.

Look at where AI gets deployed in the workflow. Eighty-nine percent of marketers use it somewhere in their influencer programs. But drill down: 36.67% use it for creator discovery, 21.11% for content generation. Then it falls off a cliff. Brief development drops to 13.89%. Reporting trails, 10.56%. And fraud detection? 7.22%.

That's the validation gap. AI is great at scanning millions of profiles fast. But the moment a contract is about to be signed and a fraudulent influencer means real reputational and legal risk, the humans step back in.

Trust ends where the audit trail begins. Brands trust AI for speed, not for stakes.

Who's actually watching, and trusting

Older audiences are still skeptical. Gen Z is not.

About 35% of Gen Z say they've bought a product promoted by a virtual personality. Seventy-five percent engage with virtual influencer profiles. And yet, only 15% rate their trust in those promoted products above a 7 out of 10. They follow. They buy. They don't fully trust.

That gap is the brand's problem to design around, not to wish away. Roughly 58% of U.S. consumers follow at least one virtual influencer. Half of brand managers who've run a virtual collaboration call the experience "very positive." And 43.8% of consumers flag ethical concerns.

So the audience is there, the engagement is there, the discomfort is also there. All three at once.

Now the rotten apples

Here's where I want to slow down.

An estimated $4.8 billion was lost to influencer fraud in 2026. That figure comes from Sumsub's deepfake and creator-economy fraud guide. And it's a sliver of a much bigger number: across all sectors, deepfake-enabled fraud caused about $23.7 billion in global losses.

Seventy-four percent of deepfake scams are run with AI tools that cost as little as $50 a campaign. Fifty dollars. Influencer-promoted scams are rising 47% year over year. And under the FTC's October 2024 rule, 2,340 creators are now under investigation by the FTC and UK FCA for fake or AI-generated reviews.

The same tech that lets one person become a media operation also lets one person become a fraud operation. The cost asymmetry is brutal: defense needs humans and lawyers, offense needs fifty bucks and a weekend.

Where this is heading

Two forces are pulling in opposite directions between now and 2030.

On one side, the generative tools keep improving and budgets keep flowing. The AI influencer market is on track for $62.67 billion. The virtual influencer high-end is forecast at $154.6 billion by 2032. Eighty percent of brand-influencer collaborations are expected to integrate AI daily by 2028.

On the other side, regulation is closing in. The EU AI Act, the FTC Endorsement Guides, the UK FCA frameworks. Industry-wide compliance costs are projected to add $200 million by 2027. AI-driven fraud losses could hit $5 billion by 2028.

The brands that come out ahead won't be the ones who went fastest. They'll be the ones who paired AI's speed with two unglamorous things: human auditing of every AI-discovered creator, and clean disclosure terms baked into every virtual-persona contract.

Back to that Calvin Klein post

I started this piece staring at a rendered face selling real clothes.

Here's what I now understand about that post. The brand behind it wasn't experimenting. They were buying three-times-higher engagement at a premium they could model. The face wasn't a person, but the contract behind it was, and the contract is where the real work happens.

The virtual creator economy isn't coming. It's here, it's $11.74 billion in size, it's growing 41% a year, and it has a $4.8 billion fraud problem sitting next to it like a bad roommate.

The opportunity is real. The trapdoor is real. The difference between the two is whether you remember to put a human in the loop before the money leaves the building.