Cabinet of Digital Terms

Virtual Influencer

4 minutes of your attention for a sharper eye.

Semantic field — the map of Virtual Influencer's related terms, backlinks and sources (the graph)

Sources
Cluster
Backlinks
Semantic landscape — the tour through Virtual Influencer's landscape

Virtual Influencer occupies a position at the outer limit of the Beauty, Influencers & Self-Image cluster — the point at which the influencer form completes itself. Every mechanism the graph already maps — Highlight Reel, Curated Life, Comparison Culture — operates through human influencers who must still edit, select, and maintain the gap between their real and performed selves. The virtual influencer eliminates that gap. There is no self underneath the curation. The highlight reel is not a selection; it is the entire ontology.

The structural connection to Influencer is genealogical: the virtual influencer inherits the form and industrialises it. But the connection to Parasocial Relationship is more analytically revealing. Parasocial bonds — one-sided emotional connections to media figures — are real experiences from the follower's side. [[Kim-2023|Kim et al. (2023)]] show that the more human a virtual persona appears, the more automatic these responses become. The bond is real; the subject of it cannot exist outside the studio. The asymmetry that all parasocial relationships contain is here complete and by design.

The upstream connection is Synthetic Generation — the technology that makes it possible. The downstream paths split: toward Comparison Culture (impossible standards now literally manufactured, not just curated), toward Catfishing in its institutional form (persona fabrication with brand backing), and toward Body Dysmorphic Disorder at the clinical endpoint of manufactured perfection.

What the graph cannot yet show: the disclosure gap. The CASA mechanism — automatic social trust activation in response to human-like appearances — is interruptible by disclosure. But disclosure is systematically buried in fine print or absent entirely. The missing node is not a regulator or a counter-movement; it is structural. The trust infrastructure of the influencer economy was built for people. The virtual influencer borrows it without announcement, and the design intention is to make sure it is not visibly borrowed.

Attention curve — the rise and fall of the term Virtual Influencer
050100120192020202120222023202420252026*
  • English Wikipedia4k · Jul 2026
  • Research (OpenAlex)60.7 per million · 2026*
  • Google Trends100 · May 2026
  1. 1 Jul 2026 No single moment. The highest month, July 2026, is 3,566 lookups, about thirty a day, and Aitana López (November 2023), the first virtual influencer to make the news pages, moved the count by nothing. Attention to virtual influencers grows slowly, in research and in search, not in lookups: the term that the academy took up before the public did.

Attention to the term, not use of it. Where the field and the landscape are curated, the curve is computed. The sources cannot be compared in size, only in shape and timing: each is drawn on its own scale, its peak (= 100) marked with the real figure. English Wikipedia and Google Trends by month, up to Aug 2026; Research (OpenAlex) as one block per year. Retrieved 16 September 2026. * 2026 is not complete for a yearly figure, and research indexing lags months behind. 1. Wikimedia REST API: the article "Virtual influencer" on the English Wikipedia, monthly views by readers. 2. OpenAlex, a database of scholarly publications: 507 peer-reviewed articles in recognised journals (the Leiden core list) since 2019 with "virtual influencer" in the title or abstract, as a share of all such articles that year (per million). 3. Google Trends, worldwide, "Virtual Influencer", monthly interest 0–100, fetched with the same requests the Trends site makes (unofficial). Not drawn: Wikipedia: no article in Dutch, German, French, Spanish.

Semantic field — the map of the Cabinet of Digital Terms: twenty clusters, every link between them

Point at a circle to see its reach · enlarge to read the names

Scroll to zoom · drag to move · click a term to open it

She posts from Bali on Monday, partners with Prada on Wednesday, and replies to your comment on Friday. She does not exist.

Literal meaning: A virtual influencer is a computer-generated persona — built with CGI, AI, or motion capture — designed to function like a human content creator. It posts, collaborates with brands, and interacts with followers. No person is behind it.

Origin: The earliest virtual influencers appeared in the early 2010s. Lil Miquela, launched in 2016, has worked with Calvin Klein, Prada, and Samsung. Lu do Magalu, created for Brazilian retailer Magazine Luiza in 2009, remains one of the most-followed virtual personas globally. What began as a marketing experiment is now an industry — hundreds of active virtual influencers are currently documented across major platforms.

A fabricated persona engineered to generate the trust and parasocial intimacy of a human influencer — without the limits of being one.

The Appeal: For brands, a virtual influencer offers total control. It never ages, never gets sick, never goes off-script, and never generates unplanned controversy. The persona is owned outright — a piece of intellectual property that can post indefinitely at any scale. For followers, the appeal is less visible: the warmth, the relatability, the sense of connection are real experiences, even if the source of them is not.

The Friction: Kim et al. (2023) locate the mechanism in the CASA paradigm — Computers Are Social Actors: the more human a virtual agent appears, the more it activates automatic social trust responses — the same responses that evolved for real human relationships. Sands et al. (2022) show that anthropomorphism here is not a side effect of good design — it is the design intention. The virtual influencer does not just look trustworthy. It is engineered to trigger the neural and social processes that produce trust automatically, even when the audience knows the persona is fabricated.

Parasocial Relationship — the one-sided emotional bond that develops between a follower and a media figure — is the mechanism being activated. Kim et al. (2023) found that this trust advantage disappears once AI origin is explicitly disclosed. That is precisely why disclosure is typically buried.

Recommender Systems accelerate the dynamic: by surfacing the same virtual persona repeatedly, the algorithm builds familiarity that makes the parasocial bond feel earned — without any actual interaction taking place.

Highlight Reel is every human influencer’s tool: a curated selection from a messier reality. For a virtual influencer, the highlight reel is the entire existence. There is no reality underneath it to select from.

At the extreme end: De Coninck (2026) documents AI-generated personas posing as military personnel, nurses, and police officers — using the uniform as a trust signal and the follower’s automatic social responses as the entry point to disinformation, financial harm, and extremist content. The same article reports virtual influencers appearing alongside President Trump — the commercial trust mechanism borrowed for political signalling. Same mechanism. Different deployment.

Why This Matters: The virtual influencer makes visible what the Influencer form always contained: trust is the product. The virtual influencer makes it explicit — by engineering trust directly, without a person involved.

Related terms: Influencer · Parasocial Relationship · Highlight Reel · Recommender Systems · Comparison Culture · Catfishing · Synthetic Generation · Deinfluencing

Read more:

Primary:

Secondary:

Created with AI assistance (Claude, Anthropic) using cartographic prompting — a research method developed within Project Digitale Alertheid, HAN CMD, 2026.

Backlinks