Guide
Strategy
Brands
A VTuber is a real person who performs live behind a 2D or 3D animated avatar, using motion capture or Live2D rigging to control the character in real time — unlike an AI or virtual influencer, which has no real person behind it at all. Ad disclosure still applies in full to the person behind the avatar, regardless of whether the audience knows their real identity, while the EU AI Act's labelling duty for AI-generated content doesn't automatically reach a live-performed avatar the same way. Major Japanese agencies like Hololive and Nijisanji typically keep the character itself when a performer leaves; some Western agencies have let the performer keep it.
A VTuber (virtual YouTuber) is a real person who performs live behind a 2D or 3D animated avatar, controlled in real time via motion capture or Live2D rigging. That's a fundamentally different construction from an AI or virtual influencer, where there's no real person behind the persona at all. The distinction isn't just theoretical: it decides who actually carries the ad-disclosure duty, and whether the EU's new rules on AI-generated content even reach the avatar.
A VTuber appears on YouTube, Twitch and other platforms behind a drawn or 3D-modelled character, while a real performer controls the character's facial expressions, movement and voice in real time — typically via face-tracking, a motion-capture suit, or Live2D rigging (a 2D image split into layers and animated to move like a character). That's neither the same as an AI or virtual influencer (no real person at all) nor a faceless or anonymous creator (a real person with no avatar, who deliberately doesn't show their face). A VTuber shows an identity — just a drawn one.
| Aspect | AI/virtual influencer | VTuber | Faceless/anonymous creator |
|---|---|---|---|
| Real person behind it? | No | Yes | Yes |
| What does the audience see? | A synthetic, computer-generated figure | A drawn or 3D avatar, performed live | No visible figure — typically hands, product or screen recording |
| Who controls the performance? | A script, pre-approved by a studio/brand | The performer themselves, in real time | The performer themselves |
| Who carries the disclosure duty? | The brand/agency behind the persona | The performer (and possibly the agency as contracting party) | The performer |
The Japanese character Kizuna AI debuted on YouTube on 29 November 2016 and is widely credited with establishing the VTuber concept itself and coining the term "Virtual YouTuber" — her channel grew from zero to over 2 million subscribers in 18 months. Major agencies have since grown up around the format: Hololive (owned by the Japanese company Cover Corp) and Nijisanji (owned by Anycolor Inc.) are today the two dominant Japanese VTuber agencies, each with dozens of characters under contract.
In the West, VShojo launched in November 2020 as one of the first US-based VTuber agencies — with a deliberately different model from the Japanese agencies, see the next section. VShojo shut down on 24 July 2025, however, after one of its own talents, Ironmouse, publicly accused management of withholding over $500,000 in charity funds raised for more than a year — a scandal that triggered a mass exodus of talent within days and the company's collapse. The example is worth knowing if a brand is considering a longer-term sponsorship through a VTuber agency: the agency's own stability and reputation is a real part of the risk picture, whatever the individual VTuber's own popularity.
The two agency models above diverge sharply on a point directly relevant to a brand considering a long-term relationship with "the face" rather than a one-off sponsorship:
| Model | Example | Who owns the design | What happens if the performer leaves |
|---|---|---|---|
| Agency-owned character | Hololive (Cover Corp), Nijisanji (Anycolor) | The agency | Cover Corp's own guidelines state the company retains its copyright and related rights — and repeated, independently reported examples show the character is typically retired permanently at "graduation," with the performer unable to reuse the model or name |
| Talent-owned character | VShojo (until its 2025 shutdown) | The performer | The performer takes the character with them — seen when Silvervale and Veibae left VShojo in 2023 and kept their own characters and assets |
The practical consequence for a brand: if you're building a long-term relationship around "the character" rather than the person behind it, confirm who actually controls it before investing in a lasting relationship — if the agency owns the character and the relationship with the current performer ends, an entirely different person could end up behind the figure going forward, or the character could disappear permanently.
Because there's always a real person behind a VTuber, the ordinary ad-disclosure duty under Markedsføringsloven § 6(4) applies in full — the duty attaches to whoever receives the commercial benefit, not to whether the audience knows that person's real identity. A VTuber performing anonymously has to disclose a paid collaboration exactly as a named influencer would — anonymity toward the audience changes nothing about the commercial intent that has to be disclosed.
Here's an important, often-overlooked distinction. The EU AI Act's Article 50 targets AI-generated content — synthetic audio, image, video or text produced by an AI system, and specifically "deepfakes" resembling an existing or plausibly existing person. A VTuber avatar controlled in real time by a performer's own facial expressions and movements via motion capture or Live2D rigging isn't itself AI-generated content in that sense — it's a performance, much like an actor in a costume, just digitally rendered.
This is an application of an existing legal framework to a new context, not a confirmed legal conclusion from a regulator or a court. The assessment could differ if the same production separately uses generative AI for part of the content — for example an AI-written script presented as the performer's own spontaneous words, or an AI-cloned voice replacing the performer's own. In that case the labelling duty would apply to that specific AI-generated element, exactly as it would for any other creator using those tools — see TikTok Symphony's avatars for a related example of how that assessment is made for an avatar built on a named person's likeness.
VTubers perform primarily via livestreaming — most commonly on YouTube and on Twitch, where the same mechanisms (Bits, subscriptions, Channel Points, sponsorships priced against concurrent viewers) apply to a VTuber exactly as to any other streamer — the avatar changes nothing about the platform's own monetization or sponsorship mechanics. On YouTube, many VTubers similarly use Super Chat and Channel Memberships, as YouTube's own Creator resources describe, to convert live viewership into direct audience support during a stream.
| VTuber | Ordinary on-camera creator | AI/virtual influencer | |
|---|---|---|---|
| Real person? | Yes, behind an avatar | Yes, shown directly | No |
| Audience connection | Parasocial toward the character, but built on a real person's personality and live presence | Direct parasocial connection to the person shown | Weaker — the audience knows it isn't human |
| Ownership risk for the brand | Can be agency- or talent-owned — see the ownership section | Rests with the person themselves, no dispute possible | Typically the brand's own, if agreed with the studio |
| Typical platform | YouTube, Twitch | Any platform | Instagram, TikTok |
| Disclosure duty | Ordinary ad disclosure; the AI Act typically doesn't reach the avatar itself | Ordinary ad disclosure | Ordinary ad disclosure plus the AI Act's Article 50 |
IF your audience overlaps with gaming, anime or livestreaming culture → a VTuber can reach an audience that doesn't necessarily distinguish a drawn character from an ordinary streamer in credibility.
IF you want a long-term, recognizable "face" independent of any one specific person → first confirm whether the character is agency- or talent-owned, see the ownership section above.
IF your contract or compliance process requires real identity verification → confirm with the agency or performer how the legal identity is documented; anonymity toward the audience doesn't mean anonymity toward the brand's contract.
IF the collaboration includes a separate AI-generated element (script, voice, clip) → assess that element on its own under the AI Act's Article 50, not the avatar as a whole.
We found no active Danish or Nordic VTuber with a documented following on the scale of the major Japanese or former Western agency characters. If a Danish-market-facing VTuber scene exists, it's by all indications hobbyist and niche today. A brand wanting to reach a Danish audience through this format should therefore expect to either work with an international, English-speaking VTuber, or accept that the immediate Denmark-specific relevance is limited — the same honest caveat already used for other niche channels on the Academy.
The figures below are entirely hypothetical, for illustration only — not a documented market rate and not a real Make Influence customer case.
A brand agrees a 2-hour sponsored segment with an independent VTuber at an illustrative rate of DKK 6,000, with a unique discount code mentioned three times during the stream. 45 viewers redeem the code with an average order value of DKK 220, producing tracked revenue of 45 × DKK 220 = DKK 9,900 and a tracked ROAS of 9,900 ÷ 6,000 = 1.65×. As with other livestreaming sponsorships, a verbally mentioned code doesn't necessarily catch every buyer, so the real result may be higher than the tracked figure — but that can't be documented without a dedicated tracking setup.
Make Influence's tracking model — a unique link or discount code per creator — doesn't care whether the creator shows their own face, an avatar, or nothing at all: the mechanic is the same. What changes with a VTuber collaboration isn't the tracking, it's the considerations covered above — who's actually the contracting party, and who owns the character if the relationship needs to last. We don't have a documented VTuber customer case of our own to point to here, so we've covered the format's opportunities and pitfalls generally, not from our own specific experience with it.
No. A VTuber is a real person performing behind an avatar in real time; a virtual/AI influencer has no real person behind it at all. See AI and virtual influencers vs human creators for that distinction.
Yes. The disclosure duty attaches to whoever receives the commercial benefit — not to whether the audience knows that person's real identity. See influencer marketing disclosure rules in Denmark and the EU.
It depends on the agency. Hololive and Nijisanji typically retain the character if the performer leaves; VShojo's talent-owned model let the performer keep the character. See the ownership section above.
Both models exist. Some VTubers are independent and negotiate directly; others are contracted to an agency like Hololive or Nijisanji, where the deal typically has to go through the agency.
Generally not to the live-performed avatar itself, but yes to any separate AI-generated element in the same production (an AI-written script, a cloned voice). See the section above.
We found none with a documented, sizeable following. The scene is, by all indications, niche or hobbyist in Denmark today.
No. A faceless creator is a real person with no visible figure at all; a VTuber shows an avatar, just not the performer's own face. See faceless and anonymous creator accounts for that distinction.
Make Influence
Find creators with real audience data, run collaborations in one place, and see clicks and sales per creator while the campaign is live.
Book a demoCreate accountMake Influence
Apply to campaigns from brands that are actively looking, follow your own clicks and sales, and get paid without chasing invoices.
Create creator profileMore creator guidesMake Influence
Briefs, agreed terms, tracking links and results sit together — so brands and creators see the same numbers.
See how it worksBrowse the Academy
