The Synthetic Star: What AI Influencers Really Cost

Overview

Virtual influencers are no longer a novelty experiment. Brands run real campaigns with them, they attract genuine audiences, and on the one metric marketers care most about they outperform human creators by a wide margin.

HypeAuditor's panel data puts virtual influencer campaigns at a 5.67% engagement rate against 1.89% for human campaigns of comparable following — roughly three times the return on the same audience size.

The commercial case is usually completed with a second claim: that synthetic personalities carry no reputational risk, because an avatar cannot misbehave. That claim has aged badly. The FTC's Endorsement Guides now expressly treat virtual influencers as endorsers, disclosure violations carry penalties above $53,000 each, and a dedicated AI enforcement unit was established in January 2026.

The risk did not disappear when the human did. It changed from reputational to regulatory, and regulatory risk is considerably easier to quantify — and considerably harder to apologise your way out of.

Why the Market-Size Numbers Don't Agree

Anyone researching this sector encounters a strange problem immediately: no two sources agree on how large it is, and the disagreement is not marginal.

Published 2026 estimates for the virtual influencer market include $1.37 billion in annual brand spending, $8.3 billion, $11.74 billion and $15.5 billion. These are not different metrics described loosely — they are presented as the same figure, and the largest is more than ten times the smallest.

Most originate from search-optimised statistics aggregators that cite each other rather than any primary survey. Growth forecasts in the same documents routinely project compound annual rates above 40%, which is a red flag rather than a finding. The reasonable conclusion is that nobody reliably knows the market's size, and that any single confident figure should be treated as marketing rather than measurement.

The One Number That Holds Up: 3x Engagement

Amid that noise, the engagement comparison is unusually well-sourced. HypeAuditor measures campaign performance across a consistent panel, and its 2026 data puts virtual influencer campaigns at 5.67% engagement against 1.89% for human campaigns.

The methodology matters here in a way it does not for the market-size claims. This compares campaigns of equivalent audience size within one measurement framework, rather than aggregating vendor self-reports, so the ratio is meaningful even if the absolute values shift.

Individual campaigns support the pattern. Prada's collaboration with Lil Miquela reportedly produced around 30% higher engagement than the brand's campaign average — a smaller premium than the panel-wide figure, and from a brand whose baseline was already strong.

Why Novelty Might Explain the Gap

A three-times engagement premium demands an explanation, and the most likely one is uncomfortable for the sector: virtual influencers are still unusual, and unusual things get looked at.

Engagement measures interaction, not persuasion. A user who pauses on an AI-generated face because it is strange, comments to ask whether it is real, or shares it as a curiosity has engaged by every analytic definition while doing nothing a brand actually wanted. Novelty-driven engagement is real engagement and poor advertising.

The test is durability. If the premium persists once synthetic personalities are commonplace, it reflects something genuine about the format. If it decays as they become ordinary, brands will have spent several years optimising for a measurement artefact — and there is not yet enough longitudinal data to say which.

What 30% Actually Was — a Projection, Not a Result

The widely repeated claim that brands have moved a large share of influencer budgets to avatars traces back to a specific and much narrower source: Ogilvy projected that chief marketing officers would allocate 30% of their influencer marketing budgets to virtual personalities by 2026.

That is a forecast about intentions, published in advance. It is not a measurement of spending that occurred, and it has been widely recycled as though it were. Against a total influencer marketing industry somewhere in the region of $31 to $40 billion, a 30% allocation would imply roughly $10 billion — which sits uneasily beside the same sources' estimate of $1.37 billion in actual annual virtual influencer spending.

That inconsistency is informative. When a projected budget share implies spending seven times higher than the observed figure, the projection is the number to discard.

What a Virtual Influencer Actually Earns

Lu do Magalu, the Brazilian retail avatar and one of the most commercially established virtual personalities anywhere, earned approximately $2.5 million in 2024 across 74 brand collaborations.

That averages to roughly $34,000 per collaboration. It is a genuinely successful commercial operation and it is not the economics of a category that has displaced human creators — top human influencers command comparable or higher fees per campaign, and there are vastly more of them.

The figure also clarifies where the savings actually come from. A virtual influencer does not undercut a human on headline fee; it undercuts on production. The same asset can be re-rendered, re-versioned and localised without rebooking talent, and that reusability is the real cost advantage rather than the sticker price.

Zero Scandal Risk Is the Wrong Frame

The strongest argument for synthetic endorsers has been that they cannot embarrass you. There are no old posts to surface, no private conduct to become public, no possibility of the face of a campaign saying something unscripted.

That is true and it is not the same as being risk-free. The FTC's updated Endorsement Guides make clear that virtual influencers, fake reviewers and anything that appears to be an individual, group or institution count as endorsers. When such a persona communicates what consumers would reasonably take to be someone's genuine opinion or experience, it is an endorsement and carries every obligation that entails.

The liability simply moved. A human influencer's scandal is their own and a brand can distance itself. A synthetic influencer is wholly brand-owned, which means every disclosure failure, exaggerated claim and misleading impression belongs unambiguously to the advertiser, with no one else to blame.

The Dual Disclosure Rule Most Brands Miss

The compliance requirement that catches campaigns is not the familiar sponsorship tag. Current FTC guidance requires two separate disclosures, not one combined label, and satisfying only the first is the common failure.

New York's Synthetic Performer Bill

Federal rules are not the only exposure. New York's Synthetic Performer Disclosure Bill passed in June and awaits the governor's signature. It would require advertisers to conspicuously disclose when an advertisement features a synthetic performer — a digitally created asset generated using AI to simulate a human likeness or performance.

The penalties are modest by comparison: $1,000 for a first offence and $5,000 for subsequent violations. Their significance is structural rather than financial, because they establish state-level obligations layered on top of federal ones.

For a national campaign that is the harder problem. A single virtual influencer appearing across every market has to satisfy the strictest applicable rule everywhere, and a patchwork of state requirements erodes precisely the frictionless scalability that made synthetic talent attractive.

What Brands Are Actually Buying

Stripped of the hype, the proposition is narrower and more defensible than the marketing suggests. A virtual influencer offers total creative control, unlimited reusable output from a single asset, straightforward localisation, and no scheduling or talent-management overhead.

What it cannot offer is the thing human influence is actually built on. A recommendation carries weight because someone with their own reputation staked it on the product. A brand-owned avatar recommending its owner's product is an advertisement wearing a face — which audiences can be perfectly happy with, provided nobody pretends otherwise.

That is why the disclosure rules matter more than the engagement statistics. The format works commercially when it is honest about what it is. The failure mode is not audience rejection of synthetic personalities; it is a regulator deciding that consumers were led to believe a person endorsed something.

Conclusion

Virtual influencers earn their engagement premium. Roughly three times the interaction rate of human campaigns is a real, consistently measured advantage, and the production economics behind it are genuine — one asset, endlessly re-rendered and localised, with no talent to rebook.

Almost everything else circulating about this sector should be treated sceptically. Market-size estimates vary by more than ten times between sources, the widely quoted budget-share figure was a projection rather than a measurement, and the engagement premium may partly reflect novelty that will not survive the format becoming ordinary.

The claim most worth retiring is that synthetic talent eliminates risk. It converts reputational risk into regulatory risk and concentrates it entirely on the advertiser. With disclosure penalties above $53,000 per violation, a dedicated FTC enforcement unit and state legislation arriving, brands treating avatars as the safe option have simply not read the newer bill.

Data Source and Attribution

US Federal Trade CommissionFTC Endorsement GuidesHypeAuditor

Engagement figures come from HypeAuditor's campaign panel data. Disclosure requirements, penalty amounts and the treatment of virtual influencers as endorsers come from the US Federal Trade Commission's Endorsement Guides and AI guidance; the Synthetic Performer Disclosure Bill details reflect New York state legislation awaiting signature. Market-size estimates circulating for this sector vary by more than an order of magnitude across published sources and are reported here as disputed rather than settled.

FactsFigs reviews, cleans, and cross-checks every source dataset before shaping it into a data story. Each visualization is created and designed in FactsFigs Design Studio — an internal tool developed and owned by FactsFigs — and is the original work of a FactsFigs author, not an AI-generated copy of any existing graphic. Individual assets within a visual may or may not be produced with AI tools, but the design of the visual itself is solely FactsFigs' own.

Figures are estimates at the time of publication, provided for information only — nothing here is legal or financial advice.

2026-07-20