Virtual influencers are running real ad campaigns for major brands, pulling 20x returns in some cases, and regulators are only now catching up to what's already everywhere.
Kenza Layli doesn't sleep, doesn't negotiate fees, and can pitch the Hyundai Kona in eight languages simultaneously. She's also not real. When Hyundai Morocco deployed this AI-generated influencer for a product launch, the campaign generated 20x ROI and ran over 2,000 chatbot conversations concurrently, results the brand described as its most successful influencer launch to date. That kind of number travels fast in marketing circles, and it has. According to market research cited by SNS Insider, the virtual influencer industry was valued at roughly $8.3 billion in 2025 and is now projected to surpass $45 billion by 2030, compounding at around 40% annually.
You've almost certainly seen these campaigns without realizing it. Shudu Gram, the first digitally created supermodel, has appeared in Fenty Beauty promotions and as part of Balmain's virtual model roster. Lil Miquela, the CGI character operated by LA-based Brud, executed an Instagram takeover for Prada during Milan Fashion Week. Samsung and BMW have run their own virtual persona campaigns. The Guardian reported in June 2026 that brands are now deploying synthetic influencers at scale while quietly omitting any disclosure that the person promoting their product doesn't exist. One in three Gen Z consumers, per the same reporting, have made a purchase based on a virtual influencer's recommendation, often without knowing the source was artificial.
Startups building the infrastructure for this market are well-funded. Synthesia, the AI video and avatar platform now used by 90% of Fortune 100 companies, raised a $200 million Series E at a $4 billion valuation in January 2026, led by Google Ventures with backing from Nvidia's venture arm NVentures, Accel, and Kleiner Perkins. AvatarOS pulled a $7 million seed round from M13 in early 2025 specifically to build AI-powered virtual influencer tooling for brands. Ogilvy projected that CMOs would direct 30% of their influencer budgets toward virtual personas by 2026, which, against an influencer marketing industry worth roughly $33 billion, puts the capital allocation somewhere north of $10 billion.
The case brands make for going synthetic is straightforward. A virtual influencer doesn't have a bad press cycle. It doesn't pull out of deals, doesn't age, doesn't require travel or a production crew, and can be localized without reshooting. Engagement metrics from virtual influencer campaigns consistently run 30% above human-influencer benchmarks, according to platform data cited by AutoFaceless. Cost-per-post is roughly half. For a CMO trying to justify a budget to a CFO, the math is not complicated.
What's more complicated is what's happening to the consumer on the other side of the feed.
The disclosure gap is closing, slowly #
The FTC's 2023 Endorsement Guides technically require disclosure when a brand deploys a virtual influencer, including labeling both the paid relationship and the synthetic nature of the persona. In January 2026, the FTC established a dedicated AI enforcement unit, and by late 2025 had already brought its first enforcement action specifically targeting undisclosed AI-generated advertising. Penalties run up to $53,088 per violation, meaning a 100-post campaign that skips disclosure could theoretically clear $5 million in FTC liability. The EU's AI Act, fully in force as of August 2026, goes further, mandating that synthetic content carry machine-readable labels that are "easily, instantly, and constantly visible."
In practice, enforcement is patchy and the rules were written reactively. The regulations that exist were shaped around human influencers who happen to use AI tools, not around fully synthetic personas designed from the ground up to pass as credible voices. There's no FTC-equivalent of a mandatory synthetic-persona registry. Brands running undisclosed AI influencer campaigns are taking a calculated bet that enforcement lags long enough to make the ROI worth it, and so far, for many of them, that bet is paying off.
The consumer cost is real. AI-synthetic fraud overtook bot-driven fraud as the costliest category of influencer fraud in 2026, accounting for an estimated $2.1 billion of a $4.8 billion total, according to figures cited by Playbella's 2026 Global Report on Synthetic Influence. Roughly 80% of Gen Z consumers say they wouldn't trust a product recommendation from a virtual persona if they knew it was synthetic. The gap between that number and the one-in-three who've bought based on such a recommendation tells you everything about how the disclosure problem actually works in practice: people act on content they've been kept in the dark about.
Brands building on this infrastructure now are placing a bet that regulatory frameworks either stay soft or that early movers can grandfather their practices before stricter rules arrive. That's not a new gamble in marketing history, and it usually ends the same way. The FTC didn't invent its influencer disclosure rules for fun. It built them after years of watching brands run native advertising, sponsored content, and paid partnerships without telling audiences the posts were paid for. Virtual influencers are that cycle running again, faster, at higher scale, with better ROI numbers to justify the delay. The reckoning, when it arrives, won't be gentle.
Also read: Kevin Warsh just killed the Fed's safety blanket and now everyone has to price their own risk • Defense tech startups have pulled in $14.6 billion this year and Silicon Valley is not turning back • John Ternus inherits an Apple where design answers to operations and plans to change that