In short. AI-generated influencers are no longer a prediction. Brands are testing them, agencies are selling them, and in a 2025 Billion Dollar Boy survey of 4,000 consumers in the US and UK, 76% said they trust virtual influencers for product recommendations, while 57% said digital twins of real creators erode trust in creator content. Canada’s Ad Standards updated its disclosure guidelines in October 2025 so that an AI influencer carries the same #Ad obligations as a human one, with a recommendation to say it’s virtual. Our position for toy, baby and family brands is simple. Use AI to make the work faster and better, say so when you do, and keep real people in front of parents and kids.
The 2024 version of this post said AI influencers would “soon” rival human creators. What happened instead is more interesting. Virtual personas carved out a lane, mostly in fashion, beauty and gaming, and the rest of the creator economy kept growing around them.
The research cuts both ways. In Billion Dollar Boy’s November 2025 study (4,000 consumers, 1,000 creators and 1,000 marketers across the US and UK), three quarters of consumers said they’d trust a virtual influencer’s product recommendation, and 85% of creators said they’d be open to building a digital twin with a brand. The same study found 62% of creators worried that virtual influencers are crowding the market, and more than half of consumers said digital twins erode trust in creator content (Billion Dollar Boy, 2025).
So the audience is curious and wary at the same time. That’s the brief.
When a creator recommends a mascara, the stakes are a wasted $30. When a creator recommends a crib, a car seat, a sleep aid or a toy for a three-year-old, a parent is trusting that person’s judgment about their child. That trust is the product we sell. It doesn’t transfer to a character that has never put a kid to bed.
Three things make this sharper in our category.
Kids are in the audience. Ad Standards’ guidelines say influencers with child audiences have to take extra care that children understand when content is advertising, with disclosure that’s attention-getting and age-appropriate (Ad Standards, Influencer Marketing Disclosure Guidelines, October 2025). A synthetic spokesperson in front of a seven-year-old is a harder case to make than one in front of a twenty-five-year-old.
The platforms are tightening. Meta’s 2026 restrictions on how teens are reached through ads push brands toward creators teens already follow and trust. We wrote about what that does to teen and tween marketing when it landed: fewer passive impressions, more competition for the hours teens are active, and a premium on creators with real standing in those communities.
Ottawa is watching. Canada’s proposed online harms law is still a proposal, and it has a long way to go through Parliament, but its direction is clear. The goal is to make platforms safer for young people, not to ban kids from the internet. For brands that sell to families, that’s a signal about where the rules are heading, not just a policy story.
Ad Standards, which administers the Canadian Code of Advertising Standards, updated its Influencer Marketing Disclosure Guidelines on October 8, 2025. Three points apply directly here.
An AI-generated influencer has to follow the same material-connection disclosure rules as a human one. If a brand paid for it, gifted it or has any other connection, the post carries #Ad. The guidelines also call it best practice to disclose that the spokesperson is virtual, with hashtags such as #VirtualInfluencer or #AIcreated.
If an image showing a product’s benefits or attributes is generated or significantly altered with AI, that should be disclosed too, for example with #MadeWithAI. Filters and AI tools don’t change the underlying rule that content has to stay truthful about what a product does.
Affiliate links, gifted products and event invitations are material connections and need disclosure on every post where they appear. (Our affiliate marketing programs are built around that.)
The platforms have their own layer. YouTube requires creators to disclose when they use AI to meaningfully alter or generate photorealistic content, and shows viewers an AI label when they do (YouTube Help). Meta and TikTok apply labels to AI-generated media as well. Disclosure isn’t optional anymore, and the agencies that treat it as a design element rather than a disclaimer are the ones that keep their clients out of trouble.
We use AI every week. We don’t use it to invent people.
Here’s the distinction in practice. In March 2026, when Air Canada’s CEO took heat for issuing an English-only apology, I built a French-language version of the statement with HeyGen in about five minutes, using the free tier, to show how simple the fix would have been. The point of the post wasn’t the tool. It was the line he wrote to go with it: own up to it right up front, say why you used it, and let the audience judge. AI made the message possible; disclosure made it acceptable.
That’s the rule we apply to creator work. AI is fine for production: translating a brief into French for Quebec creators, cutting a long-form video into clips, drafting caption variants for a creator to rewrite in their own voice, cleaning up audio. It is not fine for inventing a parent who never existed, scripting reviews the creator didn’t write, or passing off a digital twin as the real person without saying so.
We also choose creators the way you’d choose a creative partner, not a media buy. Follower counts and engagement rates get a creator onto the shortlist. What gets them the brief is whether they can educate and entertain, whether they’ve earned trust in a specific community, and whether they can make a brand culturally relevant across social, retail, PR and events. A virtual influencer can hit the first two numbers. It can’t do the rest for a family brand, because it has never been one.
The 2024 list had ten tells. Most of them (perfect images, a steady posting schedule, consistent quality) now describe plenty of real, professional creators. Here is the shorter list we use when vetting.
Look for a life outside the feed. Real creators get tagged by other people, show up in other creators’ content, attend things and have a history that predates their growth curve. A profile that only ever appears in its own posts is a flag.
Check the hands, the backgrounds and the continuity. Generated images still slip on fingers, text in the background, jewelry that changes between frames and rooms that don’t match from post to post.
Read the comments, not the counts. Hundreds of generic compliments in the first hour with no replies from the creator is a pattern. Real audiences argue, ask questions and get answered.
Ask for the raw files. A creator can send you an unedited clip from their phone. A persona can’t.
Ask directly. Under the 2025 guidelines, a virtual influencer should already be telling you. If a management company dodges the question, that’s your answer.
Decide your position before a vendor pitches you one. Write down where AI is welcome in your creator program (production, translation, editing, research) and where it isn’t (synthetic spokespeople, fabricated reviews, undisclosed twins). Most of the risk comes from not having decided.
Put disclosure in the brief. #Ad, #Gifted, #affiliate and, where it applies, #AIcreated, in the safe zone of the post, on every platform. Make the creator’s compliance part of the contract.
Vet creators like you’d vet a hire. The checklist above takes twenty minutes per creator and has saved us from more than one bad shortlist.
Keep real people in front of parents. If your brand is bought by someone deciding what’s safe for their child, the person recommending it should be someone who has made that decision too.
For a toy, baby or family brand, we'd push back, and we'd want to know what the persona can do that a real parent creator can't. If a client has a specific, disclosed use (a brand character with a long history, say), that's a different conversation.
If the image shows the product's benefits or attributes and it was generated or significantly altered with AI, Ad Standards says yes. Light colour correction isn't the issue; a product render presented as a photo is.
Yes. The guidelines are explicit that AI-generated influencers follow the same material-connection rules, plus the recommended virtual-influencer disclosure.
The published research says Gen Z is more curious than older groups, and the Billion Dollar Boy numbers on trust are higher than most people expect. Curiosity isn't the same as buying a car seat on a character's say-so, which is why we treat the category as the deciding factor.
About the author
Andrew Wagar founded Swerve Strategic Marketing in 2010 and runs it as President and CEO. He’s spent more than 30 years in PR and marketing, including PR Manager roles at Spin Master and the Canadian Standards Association.
He’s an Accredited Public Relations Professional (APR) with the Canadian Public Relations Society and has served on the Canadian Toy Association board since 2018, where he chairs marketing communications. Reporters covering the toy business call him when they need a source.
