Some links below are affiliate links. If you buy through them we may earn a commission at no extra cost to you. We only list tools we would use ourselves. How we make money.
An AI digital twin of yourself is an AI model trained on your own photos that puts your face into new images and videos without a camera shoot. Creators build one by uploading 20 or more clear photos to a character tool such as Higgsfield Soul ID. They generate stills of the trained face, animate the best ones into short clips and label the results as AI. Setup takes an afternoon. The harder part is deciding what your twin is allowed to say.
That question is now a business. Guardian Australia reported on 25 October that Socceroos footballer Cristian Volpato has licensed his "synthetic twin" to a California company that sells it to brands. Below: what he signed, a step-by-step build for your own twin, honest UGC ads, and the labelling rules.
What did footballer Cristian Volpato actually license?
According to Guardian Australia, Volpato is 22 and plays for Serie A club Sassuolo. In mid-September he signed with Callandor, a California-based company that acts as an AI licensing agent for professional athletes. Callandor got exclusive rights to his digital likeness, voice, biometric profiles and performance data. It will look for brand deals for his AI image, take a cut, and help restrict content he hasn't approved.
Callandor co-founder Michael Fisk told the Guardian that Volpato is always training or playing, so he misses brand placements. "A synthetic twin could be used 24-7, year-round by brands. Assuming it's approved, right?" Callandor's own press release says one verified capture lets him run official digital twins in multilingual brand campaigns. That is the company's claim; the Guardian notes that Callandor's campaigns are still in a pilot phase.
He isn't the first. The same report says Neymar licensed his AI likeness in June to a Chinese-owned entertainment platform that made short videos during the FIFA World Cup. It's also clear about what nobody knows yet: whether brands will pay as much for an AI clone as for the real person.
What does an AI twin deal mean for creators who aren't famous?
Two things, and only one of them is about money.
First, the value of a licensing deal comes from the audience, not the model. Brands want Volpato's twin because they want Volpato. A twin of a creator with 800 followers is a production tool, not a licensing asset. Still useful: it makes the shots you don't have time to film.
Second, the part worth copying is consent on the record. William Saunders, who runs the Sydney start-up Hyperknown, told the Guardian: "AI can make content abundant. It can't manufacture someone's permission." Volpato's deal is built around a record of what he has approved. The free version: a written list of what your twin may do, and a label on everything it makes.
How do you create an AI digital twin of yourself, step by step?
This walkthrough uses Higgsfield Soul ID because it trains on your real photos and carries the same face into both images and video in one account. The steps come from Higgsfield's help center and product guides. It's a research summary, not a test report; check the current price in the app.
Step 1: Write your twin's rules before you train it
Write down:
- Allowed: the content your twin may appear in, such as outfits, travel, product demos and explainers.
- Off-limits: topics you would never post about, such as politics, religion, health claims or anything involving children. Fisk told the Guardian you have to be "very, very careful" with "anything with religion and drugs".
- Products: only items you have actually used. The UGC section below explains why.
- Labels: where and how you will mark AI content: the platform's AI toggle plus a line in the caption.
Step 2: Shoot your training photos
Higgsfield's help center asks for 20 or more photos of the same person, with up to 80 supported. Its tips:
- Well-lit, high-quality photos showing your face from varied angles and with varied expressions.
- At least one full-height photo, not only headshots.
- Recent pictures, ideally from the last four to five months, so the model learns your current hair and face.
- No sunglasses, hats, scarves, heavy shadows, cropped faces or group shots.
Higgsfield also says a smaller set of clean, varied photos beats a large set of inconsistent ones. One afternoon by a big window with three outfits and a friend holding the phone is enough.
Step 3: Train your Soul ID
- On higgsfield.ai, open Soul and choose Train new character.
- Upload your photos and start training. The help center says it takes a few minutes.
- Name the character, then select it in the Character tab whenever you generate.
Higgsfield's consent rule is short: only upload photos of yourself, or of someone who has given you permission to use their likeness. One Soul ID represents one person. The identity itself can't be exported as a file; the images and videos you make are normal downloads.
Step 4: Build a reference pack of stills
Before you make any video, generate 15 to 20 images of your twin in the situations you actually post: at your desk, outdoors, holding a product, in your usual clothes. Keep the ones that look like you; they become first frames for video. A starting prompt, with your Soul ID selected:
Candid phone-camera photo of the selected character sitting at a sunlit kitchen table, holding a coffee mug, mid-laugh while looking at the camera, soft morning window light, natural skin texture, slight grain, vertical 9:16
Expect the same person, not a pixel-identical face. Higgsfield warns that weak photos, extreme style shifts or unusual angles can cause drift, so start with realistic everyday scenes before you try a fantasy look. If you need help with wording, draft prompts in our free prompt studio.
Step 5: Turn your twin into video
Higgsfield offers two routes:
- Motion clips in Seedance. Trained Soul ID characters show up in Seedance automatically as an Element. Type
@in the prompt (or click @ Elements), pick yourself from My Elements, then describe the action: walking into frame, opening a parcel, turning to the camera. - Talking clips in Lipsync Studio. Start from one of your stills and add a script or an audio clip. Higgsfield's guide says that when the source image comes from Soul ID, the trained identity carries into every generation. Put emotion cues in brackets, such as [excited] or [calm], keep each generation under 30 seconds and cut between clips, because longer takes start repeating gestures.
Lipsync Studio animates a still image to audio; Higgsfield's guide says it doesn't do video-to-video dubbing. For the cleanest mouth movement, use a sharp, evenly lit, front-facing still without glasses.
Step 6: Check every output before it goes out
Zoom in on hands, teeth, jewellery and any text on products. Would you be comfortable if a follower learned exactly how it was made? If not, regenerate or film it for real.
How do you use your AI twin for UGC ads?
In Higgsfield's Marketing Studio, you upload a product image or paste a product page link, pick a template and generate. The UGC Videos category has Faceless, Talking Head and Silent formats. Higgsfield's help center suggests training a Soul ID once so every variation uses the same face. Clips run up to 15 seconds per generation, and Higgsfield notes that voiceover and pacing can vary between generations, so review each one before publishing.
ImagineArt is the simpler option if you only need a talking head. Its AI talking avatar tool takes one photo and a typed script and returns a lip-synced video with expressions and hand gestures, powered by HeyGen Avatar IV. ImagineArt says generations run up to 3 minutes, credits are charged per second of avatar video, scripts work in many languages, and you don't need a HeyGen account. For stills, its AI Influencer app builds a consistent visual persona for feed posts.
| Task | Higgsfield | ImagineArt |
|---|---|---|
| Learn your face | Soul ID, trained on 20 to 80 photos | Talking avatar from a single photo |
| Consistent stills | Soul image models with your character selected | AI Influencer app for a consistent persona |
| Motion video | Seedance, with your twin as an @Element | Image-to-video with models such as Kling and Seedance |
| Talking video | Lipsync Studio, under 30 seconds per take recommended | HeyGen Avatar IV, up to 3 minutes per generation |
| Product ads | Marketing Studio UGC: Faceless, Talking Head, Silent (up to 15 s) | Spokesperson and UGC-style ads from a photo and script |
The honesty rule for ads. Under the US FTC's Endorsement Guides, endorsements must reflect the honest opinions, findings, beliefs or experience of the endorser. Your twin is still you, so it shouldn't say "I've used this every day for a month" about a product you have never opened. Paid posts also need a clear ad disclosure such as #ad, on top of any AI label. If a brand wants to use your twin, put the terms in writing: which channels, for how long, who approves each video, and what happens to the files when the deal ends.
Do you have to label AI content made with your own face?
In most cases, yes: a realistic video of you saying a script you never filmed is what these rules describe.
- TikTok requires creators to label AI-generated or significantly edited content that shows realistic-looking scenes or people. Unlabelled content may be removed, restricted or labelled by TikTok.
- YouTube requires disclosure when realistic AI content makes a real person appear to say or do something they didn't do, or shows a realistic scene that didn't happen. Cloning your own voice for voice-overs or dubs is listed as not needing disclosure, but a talking twin in a scene you never filmed falls under the first rule. YouTube says disclosure alone won't limit a video's audience or its eligibility to earn money, while not disclosing can lead to penalties, including suspension from the YouTube Partner Program.
- Instagram, Facebook and Threads: Meta requires people to use its disclosure and label tool when they post photorealistic video or realistic-sounding audio that was digitally created or altered, and says it may apply penalties if they don't.
- European Union: since 2 August 2026, Article 50 of the AI Act requires anyone who uses AI to publish a deepfake image, audio or video to disclose that it was artificially generated or manipulated. For evidently artistic, satirical or fictional work, a lighter disclosure applies.
One habit covers all of these: switch on the platform's AI label and add a plain line to the caption, such as "Made with my AI twin."
What are the risks of making an AI twin of yourself?
- Other people using your face. The Guardian lists cases from sport: an offshore casino used a deepfake of Manchester United's Bruno Fernandes in advertising, and the AFL told its betting partner Sportsbet to stop using AI images of players without consent. Report impersonation through each platform's tools.
- Intimate deepfakes. In the US, the TAKE IT DOWN Act requires covered platforms to remove nonconsensual intimate images, including AI-generated ones, within 48 hours of a request. If a platform doesn't act, the FTC takes reports at TakeItDown.ftc.gov.
- Account access. Your trained character lives in your account, and anyone who can log in can generate you. Use a unique password and two-factor login where offered.
- Near-misses. Fisk told the Guardian the difference is "becoming almost impossible to detect", particularly on a phone. Your followers know your face, though, so delete anything that only looks 90% like you.
- Signing away rights. Volpato's deal gave Callandor exclusive rights to his likeness. If anyone offers to license your twin, find out exactly what "exclusive" covers and for how long, and get legal advice before you sign.
Is an AI twin worth it for a small creator?
If you post on a schedule and filming is your bottleneck, it can be, as a production tool. A twin can cover in-between posts and test ad hooks while you film the posts that need the real you. If you hope a twin alone will bring brand deals, it won't: brands pay for audiences. Build the audience with your real face, and use the twin to post more consistently.
Prefer not to use your face at all? A fictional AI character works with the same tools. Start with our guide to making money with an AI influencer and these AI influencer niche ideas, or try this brief:
An original virtual influencer, a confident woman in her mid-twenties with a copper-auburn bob, light freckles and small gold hoops, wearing an oversized sage linen shirt, relaxing on a terracotta rooftop terrace above a hillside coastal city and holding an iced matcha, caught mid-smile as if a friend just took the shot. · Photorealistic 35mm Film Frame · Camera: Full-frame mirrorless, 35mm prime at f/1.8, 1/500s, ISO 100, slight handheld tilt for a candid feed look · Lighting: Setting sun as warm key from behind-left, soft bounce from white walls filling the face, gentle lens flare at frame edge · consistent character sheet ready, same face across shots, natural skin texture with pores, subtle Portra 400 grain, lifestyle Instagram aesthetic, shallow depth of field, authentic candid pose --ar 4:5 --no plastic skin, airbrushed, extra fingers, distorted hands, celebrity likeness, brand logos, text, watermark, oversaturated, revealing clothingCustomise it in the free studio
A sensible first week: shoot your training set, train one Soul ID, make five labelled posts, and only then decide whether it actually saves you time. For more ways to put these tools to work, see AI side hustles for 2026 or browse every tool in our tool index.
FAQ
What is an AI digital twin of yourself?
An AI model trained on photos of you, and sometimes your voice, that generates new images and videos with your face. Creators use one for content and ads; athletes such as Cristian Volpato license theirs to brands.
How many photos do I need to make an AI twin?
Higgsfield Soul ID asks for at least 20 photos of the same person, up to 80, with varied angles, good light and at least one full-height shot. ImagineArt's talking avatar works from a single photo.
Can I make an AI twin of someone else?
Only with their permission. Higgsfield tells users to upload only photos of themselves or of someone who has agreed to the use of their likeness, and platforms require disclosure when AI makes a real person appear to say or do something.
Do I have to label content made with my AI twin?
For realistic content, yes. TikTok, YouTube and Meta all require AI disclosure for realistic people or scenes, and in the EU deepfake disclosure has applied since 2 August 2026. Paid posts also need an ad label such as #ad.
Can I make money licensing my AI twin?
Only if brands already want you. Volpato's deal works because he is a pro footballer, and even his campaigns are pilots. Most beginners use an AI twin to save shooting time, and most new accounts earn little or nothing at first.
Sources
- Guardian Australia: 'Synthetic twin': Why Socceroo Cristian Volpato is the latest footballer to license his AI clone
- Callandor press release (Business Wire via Medianet): Football star Cristian Volpato partners with Callandor to take control of his AI identity
- Higgsfield Help Center: How do I create and use a Soul ID character?
- Higgsfield Help Center: How do I use Marketing Studio to create video ads?
- Higgsfield: Lipsync Studio, turn any script into a talking performance
- ImagineArt: Free AI Talking Avatar Generator
- ImagineArt: AI UGC Video Generator
- Federal Register: FTC Guides Concerning the Use of Endorsements and Testimonials in Advertising (2023)
- TikTok Community Guidelines: Integrity and authenticity
- YouTube Help: Disclosing use of altered or synthetic content
- Meta: Labeling AI-generated images on Facebook, Instagram and Threads
- EU AI Act, Article 50: Transparency obligations
- FTC Consumer Advice: FTC enforces compliance with the TAKE IT DOWN Act