This guide to history of virtual influencers covers definition, examples, and why it matters.
The Prehistory: Before Anyone Called Them Influencers
The story of virtual influencers does not begin on Instagram. It begins on stages, screens, and album covers in the 1980s and 1990s, decades before “influencer” was a word and decades before any social platform existed for them to live on. To understand why the category exists today, it helps to understand that the cultural appetite for fictional digital personalities is older than the internet.
In 1981, the British synth-pop group The Buggles released a music video featuring an animated singer. In 1985, Max Headroom, a CGI-styled (actually rubber-mask) character debuted as a host on Channel 4 in the UK and quickly became a global pop-culture phenomenon, doing TV interviews, hosting talk shows, and starring in commercials. Max Headroom was arguably the first “virtual celebrity” in the modern sense, a fictional digital character with a personality, a fan base, and brand deals. He was not a virtual influencer because Instagram did not exist yet, but the template was set.
Through the 1990s, anime and visual novel cultures in Japan developed an entire economy around fictional characters that fans formed real attachments to. Yuki Terai, a Japanese 3D virtual model who appeared on TV and in print campaigns starting in 1996, was an early commercial application. By the early 2000s, brand mascots like Lu do Magalu (launched by Brazilian retailer Magazine Luiza in 2003) demonstrated that a virtual character could be a long-term brand asset, even though the term “virtual influencer” did not yet exist.
2007: Hatsune Miku and the First Mass-Audience Virtual Star
The big leap came in 2007 when Crypton Future Media in Japan launched a vocaloid voice synthesis software character named Hatsune Miku. She was originally a piece of software, a virtual singer that musicians could use to generate vocal tracks, but she rapidly took on a life of her own. Fans created songs for her. Animators created videos. Game developers built her into rhythm games. Within a few years she was performing sold-out concerts as a hologram on stage backed by a live band, in Tokyo, Los Angeles, Singapore, and dozens of other cities.
Miku is not technically a “virtual influencer” in the social-media sense, but she proved a critical point: an audience would form deep emotional and financial attachments to a fictional digital being. People would buy her merchandise, attend her concerts, write fan fiction about her, and treat her as a real cultural figure. The economic and emotional infrastructure for the modern virtual influencer category was being built around Miku for almost a decade before Lil Miquela ever posted her first photo.
2016: Lil Miquela and the Birth of the Modern Category
In April 2016, an Instagram account named @lilmiquela posted its first photo. The account belonged to a CGI character named Miquela Sousa, supposedly a 19-year-old Brazilian-American woman living in Los Angeles. The photos looked unusual, slightly off, slightly too perfect, but plausible enough that many followers initially could not tell whether she was real. Speculation about her identity drove early growth.
Behind Miquela was a Los Angeles startup called Brud, founded by Trevor McFedries (a music industry veteran) and Sara DeCou. Their bet was that the existing influencer economy, just hitting its stride in 2016, would eventually want characters that brands could fully control, and that Instagram was ready to host them. They were right. By 2017 Miquela had hundreds of thousands of followers and her first brand collaborations. By 2018 she had a Calvin Klein campaign, a feature in Time magazine’s “25 Most Influential People on the Internet,” and several million followers. The category as we know it today was born.
What made Miquela different from Hatsune Miku and the earlier virtual celebrities was that she was operated like an influencer: posting daily-life photos, sharing emotions, partnering with brands the way any human creator did. She was not a singer or a mascot. She was a personality who happened to be rendered.
2018 to 2020: The Wave
Miquela’s success kicked off an explosion of imitators and competitors. Imma launched in Japan in 2018, created by Tokyo studio ModelingCafe and operated by Aww Inc., with a photoreal style that pushed the visual quality of the category forward. Shudu, “the world’s first digital supermodel,” launched in 2017 (predating but not commercially comparable to Miquela’s peak) and fronted campaigns for Balmain and BMW. Noonoouri, a stylized French character, started doing fashion editorials with luxury brands. Knox Frost (also from Brud) became one of the first prominent male virtual influencers.
This period also saw the first serious controversies. Shudu’s case became a flashpoint when it emerged that she was created by a white photographer, raising questions about race, representation, and who profits from the aesthetics of communities they do not belong to. Lil Miquela’s “kidnapping by Trump supporters” stunt in 2018, a staged narrative that some interpreted as exploitative of real political tensions, drew criticism from cultural commentators. The category was generating real debate in mainstream media for the first time.
Brand budgets started flowing in seriously. Reports estimated that the global virtual influencer marketing industry was worth a few hundred million dollars annually by 2019, growing rapidly. Major fashion, beauty, and luxury automotive brands all ran their first virtual influencer campaigns in this window.
2020 to 2022: Asia Takes the Lead
While the West kept its attention on Lil Miquela and a small cluster of LA-based characters, the most rapid expansion of the virtual influencer category was happening in Asia. South Korea launched Rozy in 2020, who crossed one million dollars in earnings in her first year. Japan’s Imma went mainstream with deals from Ikea, Porsche. Coach. Chinese tech platforms like Bilibili and Douyin built features specifically to support virtual personas. Brazil’s Lu do Magalu kept growing into the most-followed virtual character in the world by raw count.
By 2022 the virtual influencer category was no longer a Western-led phenomenon. The most active markets were Japan, South Korea, China. Brazil, with Western markets representing a smaller share of total revenue and a smaller share of total characters in operation.
This period also saw the first generation of audiences who had grown up with virtual celebrities and treated them as natural rather than novel. The cultural friction that surrounded early Western virtual influencers, “is this weird?”, “are people really following CGI characters?”, was largely absent in markets that had grown up with vocaloid stars and anime culture.
2022 to 2023: Generative AI Breaks the Door Open
Then ChatGPT launched in November 2022, Stable Diffusion released its first widely usable version in August 2022, Midjourney v4 came out in late 2022. DALL-E 2 launched mid-2022. In a span of eight months, the cost of producing high-quality character images dropped by roughly two orders of magnitude. Simultaneously, the cost of producing in-character text dropped by roughly the same factor.
Before this period, building a virtual influencer required either a 3D modeling studio (slow and expensive) or a small team of artists, writers, and producers (also slow and expensive). After this period, a single person with a laptop could produce dozens of consistent character images per day and write captions in a custom voice using a chatbot. The economic structure of the category was rewritten in less than a year.
The first wave of generative-AI-built virtual influencers started appearing in mid-2023. Most were rough; the technology was still maturing and creators were figuring out the workflows. By late 2023 the quality was good enough that some of these new characters were competing with studio-built ones for audience attention and brand budgets.
2024: The Long Tail Explodes
By 2024 the long tail of virtual influencers was the most dynamic part of the category. Solo creators, small teams, and content studios were launching new characters at a pace the studio era could not have matched. Tools like Vinfluencer.ai started appearing to give long-tail creators the chat, monetization, and persona-management infrastructure that previously only top studios had.
Aitana Lopez launched in Barcelona in mid-2023 and reached commercial viability within months, becoming a template for what an independently built virtual influencer could achieve in the new generative era. Hundreds of similar characters followed in 2024 across Spain, Mexico, the Philippines, Indonesia, Vietnam, and other markets where the lower production costs made commercially viable virtual influencer projects possible for the first time.
The category also saw its first wave of consolidation. Several early studios were acquired by larger entertainment companies. Brud was integrated into a larger entertainment holding. Specialist agencies emerged to broker between long-tail characters and brand partners.
2025 to 2026: Maturity and the Next Frontiers
The current state of the category is one of maturity. Virtual influencer marketing is a routine line item in major brand budgets. The top characters earn millions of dollars a year. The long tail has tens of thousands of operators, with the top decile crossing into commercially viable revenue. Cultural acceptance is high in most markets and the “is this weird” debates have largely faded.
What is changing now: real-time video generation is becoming viable, voice synthesis is closing the gap with human voices, autonomous AI agents are starting to take over the operational side of virtual influencer management, and the line between virtual influencers and AI companions is blurring as both categories converge on similar technology stacks.
The next frontier, expected over the next 24 months, is virtual influencers that operate semi-autonomously, with AI agents handling day-to-day decisions about what to post and how to engage with fans, and human operators stepping in only for strategic direction and brand-deal management. When that happens, the long tail will explode further and the production cost of running a virtual influencer will drop close to the cost of running the underlying language model.
What the History Tells Us
A few patterns are worth pulling out of the timeline.
First, virtual influencers did not arrive suddenly. They emerged from a 30-year tradition of fictional digital personalities, from Max Headroom through Hatsune Miku to Lil Miquela. The cultural infrastructure for accepting them as “real” enough to follow was built decades before the platforms existed.
Second, the category has consistently been bigger and faster-moving in Asia than in the West. Western media coverage centers on Lil Miquela because she is the most familiar to Western audiences, but the actual scale of the category is dominated by Japanese, Korean, Chinese. Brazilian characters and audiences.
Third, every major step-change in the category has come from a technology shift. The first wave came from 3D modeling and Instagram. The second wave came from generative image models. The next wave will come from real-time video and autonomous agents. The pattern is consistent and the next shift is already visible on the horizon.
Fourth, the category’s controversies are not new and the criticisms are not getting louder over time, they are getting more sophisticated. Early concerns about authenticity have given way to more substantive debates about beauty standards, race, and the cultural effects of normalizing relationships with non-human personas. These debates will continue and will shape the regulatory environment over the next few years.
Conclusion
The history of virtual influencers is shorter than people expect (the category is roughly a decade old in its modern form) and longer than people expect (the cultural lineage stretches back 30+ years). It is a story of how technology, culture, and commerce converged to create a new category of media that did not exist five years ago and that millions of people now interact with daily.
If you are building in this space, the history is useful because it shows where the next opportunities are. The long tail is bigger than the top of the market. The next technology shift will reshape costs and capabilities again. The Asian markets are larger than the Western ones. The cultural debates will keep evolving.
Vinfluencer.ai exists in the current chapter of this story, the era when individual creators have access to the same tools that previously required studios. The next chapter will be written by the people building virtual influencers right now.
Related reading on Vinfluencer
Further reading
Further reading
You must be logged in to post a comment.