Fake creator endorsements are spreading across social platforms, putting influencer credibility, brand safety and consumer trust under growing pressure.
Deepfake influencer ads are spreading across social media, threatening creator trust, brand safety and the future of influencer marketing.
Introduction
The latest social-media problem is not another dance challenge or trending sound. It is something far more consequential: AI-generated versions of real creators appearing to endorse products they never approved.
The issue is moving quickly from an obscure corner of artificial-intelligence culture into the mainstream creator economy. A fresh investigation published by The Guardian has highlighted how influencers are discovering fake advertisements using digitally altered versions of their faces to promote products and services they have never worked with. Lifestyle creator Emily Schuman, for example, found advertisements using manipulated versions of her likeness to promote companies with which she had no relationship.
The wider problem is appearing in multiple countries and sectors. Australian regulators are warning about deepfake celebrity endorsements used in investment scams, while researchers have uncovered networks using AI-generated videos of politicians across Australia and other Western countries.
For creators, brands and audiences, the stakes are becoming clear: if a face and voice can be copied convincingly, what happens to trust?
What is the deepfake influencer trend
A deepfake is synthetic media created or manipulated using artificial intelligence to make a person appear to say or do something that never happened.
In influencer marketing, the technology can be used to:
Alter a creator's face in an existing video.
Generate an artificial version of a creator holding a product.
Clone or imitate a creator's voice.
Change what a person appears to say.
Produce completely fabricated endorsements.
Redirect audiences toward fraudulent websites or products.
There is an important distinction between legitimate virtual influencers and unauthorised deepfakes.
A brand can create a fictional AI personality and openly tell its audience that the character is synthetic. The controversy begins when a real person's identity is copied without permission.
That distinction matters because an influencer's face is not merely an image. It is part of a commercial reputation built over years.
Alice Marwick of the Data & Society Research Institute, who has studied AI impersonation scams, told The Guardian that an influencer's brand is effectively their business. If that brand is damaged, the consequences can reach their wider professional life.
How the trend started
The use of fake celebrity endorsements predates generative AI. For years, scammers have manipulated photographs, edited videos and created fake websites to make famous people appear to recommend financial products or other services.
Generative AI has dramatically lowered the technical barrier.
Instead of needing advanced video-editing expertise, bad actors can increasingly create convincing images, voices and videos using readily available AI systems.
Australia provides one of the clearest current examples.
The Australian Securities and Investments Commission, or ASIC, recently warned consumers that scammers are using deepfake videos of celebrities, politicians and financial commentators to promote fake investment opportunities. ASIC said it removed more than 19,400 online scams in the relevant financial year, an increase of 182% from the previous year.
The regulator says the fake endorsements are often only one piece of a larger deception involving cloned news websites, fabricated reviews and fake investment platforms.
That makes deepfake influencer content more than a technology story. It is becoming an infrastructure for online manipulation.
Why deepfake influencer content is spreading so quickly
Familiar faces create instant credibility
Influencer marketing works because audiences recognise the person delivering the recommendation.
A creator may have spent years establishing a particular identity: the beauty expert, the fitness coach, the financial commentator, the fashion personality or the lifestyle blogger.
A fake advertisement can exploit that accumulated trust instantly.
A viewer who sees a familiar creator holding a product may assume there is a genuine sponsorship relationship.
That is particularly dangerous when the advertisement is designed to look like ordinary social content rather than a traditional commercial.
AI is becoming more convincing
The technology is also improving.
Some AI-generated videos still contain obvious clues — unnatural movement, strange expressions, inconsistent lighting or imperfect lip synchronisation. But sophisticated synthetic media can be much harder to identify at a glance.
Research into synthetic media is increasingly focused on the psychological consequences of this realism. Academic work published in Frontiers in Communication has examined how AI-generated “experts” can reproduce visual and vocal cues associated with authority and credibility.
The result is a difficult problem for social-media users: people are being asked to verify content using their eyes and ears at exactly the moment those senses can no longer be trusted completely.
Social algorithms reward attention
A deepfake does not have to fool every person who sees it.
It only needs to generate enough clicks, comments, shares or purchases to become worthwhile for whoever created it.
That is especially important on recommendation-driven platforms, where content can reach people who do not follow the original creator.
Researchers investigating an AI-generated influence network targeting Australian politicians recently found more than 200 Facebook pages involved in impersonation activity. The same network reportedly targeted audiences in the UK, US, Canada and other countries.
The researchers described a commercially motivated system in which content was produced, tested and monetised according to what attracted attention.
Platform influence: Instagram, Facebook, TikTok and YouTube
The major platforms are responding, but the challenge is enormous.
Instagram and Facebook
Meta has introduced AI transparency measures and has been developing systems intended to identify and label AI-generated content.
The company has also said it removed more than 20 million accounts impersonating large content creators in 2025 and has been testing tools designed to help creators detect potential impersonation.
Yet recent cases show the gap between policy and enforcement.
Schuman reported that she and her followers repeatedly reported advertisements using her likeness. One advertisement was eventually removed after weeks of complaints, while another allegedly remained in circulation.
That creates a difficult credibility problem for platforms: announcing an anti-deepfake policy is considerably easier than enforcing it at internet scale.
TikTok
TikTok has rules covering deceptive AI-generated content and identity manipulation, but the platform faces the same basic problem as its competitors.
Short-form video is designed to be consumed rapidly. Users may watch, react and scroll before they have time to determine whether a person in a video actually authorised it.
The danger becomes even greater when an impersonation is connected to a scam, because the fraudulent content can be distributed to highly targeted audiences.
YouTube
YouTube has taken a particularly notable step by developing Likeness Detection, a system intended to help eligible creators identify videos in which their face has been digitally generated or altered.
The platform has expanded the programme beyond an initial group of public-interest figures to include celebrities and talent agencies.
YouTube also requires creators to disclose realistic synthetic or altered content in situations where a real person appears to say or do something that did not happen.
That represents a broader change in thinking: platforms are increasingly treating digital identity as something that needs active protection.
Creator influence: when your face becomes a business asset
Creators have traditionally worried about copying in the form of stolen videos, reposted photographs and fake accounts.
Deepfakes introduce a more serious version of the problem.
Someone can now take material that a creator deliberately published and use it as raw material for something the creator never intended.
Fashion creator Molly Tranchin, known as FashionVeggie, has taken legal action against underwear company Eby over an alleged AI-altered promotional video. According to reporting on the case, Tranchin alleged that a video featuring her likeness was manipulated into a different pose and appearance from the content she had agreed to provide.
The legal dispute is significant because it raises a question that will become increasingly important:
Who controls the commercial use of a creator's digital likeness?
As AI improves, that question could become as important to influencer contracts as payment, exclusivity and usage rights.
Public reaction is shifting from curiosity to suspicion
People have generally shown fascination with AI-generated personalities and synthetic video.
But there is a major difference between enjoying obviously fictional AI content and being deceived by an artificial endorsement.
Research into AI-generated influencer advertising has found that disclosure and perceptions of artificiality can affect consumer responses, including brand credibility and purchase intentions.
The wider concern is trust fatigue.
If users repeatedly encounter fake celebrities, fake influencers, fake experts and fake news footage, they may eventually begin questioning authentic material too.
That creates a strange consequence of generative AI: the technology may not only make fake content more believable. It may also make real content less trusted.
Business and marketing impact
For marketers, the deepfake boom is creating two opposing opportunities.
On one side, synthetic influencers can be cheaper, scalable and highly customisable. Brands can create fictional digital personalities without negotiating traditional creator contracts.
On the other side, unauthorised use of real people's identities creates serious brand-safety and reputational risks.
A legitimate AI campaign should establish:
Explicit consent for likeness and voice usage.
Clear disclosure when synthetic media is used.
Written limits on how a creator's digital identity can be altered.
Monitoring for unauthorised copies after campaigns launch.
Rapid procedures for reporting fake advertisements.
Verification of agencies and third-party advertising partners.
The UK advertising regulator has also stressed that existing advertising rules continue to apply when AI is involved. Artificial intelligence does not provide advertisers with a special exemption from established advertising standards.
The commercial lesson is straightforward: AI can reduce production costs, but it cannot remove the need for consent.
Advantages and criticisms
There are genuine benefits to AI-generated creator technology.
Creators could eventually use authorised digital versions of themselves to produce multilingual videos, create routine content, appear in different formats and reduce production workloads.
YouTube's development of tools allowing creators to manage their own likeness points toward a future in which AI replicas could become useful extensions of a creator's business — provided the creator controls them.
The criticism is that the same technology can turn identity theft into a scalable business model.
Another concern is that AI labels alone may not solve the problem. A small “AI-generated” notice may not undo the persuasive effect of a familiar face, particularly when users encounter content while scrolling quickly.
Platforms therefore face pressure to improve both detection and enforcement, rather than relying entirely on audiences to identify fakes themselves.
Real examples show the problem is expanding
The current cases span multiple categories.
In the UK, fake advertisements have recently misused brands and celebrity identities to promote unauthorised online gambling offers. Tesco reported a scam involving its branding and fake celebrity endorsements to City of London Police.
In Australia, ASIC says celebrity and politician impersonations are being used to lure people into investment scams, often with fake news articles and fabricated testimonials supporting the deception.
Researchers have also identified AI-generated political impersonation networks operating across several Western countries, including Australia, the UK, US and Canada.
Meanwhile, the recent experience of creators such as Emily Schuman demonstrates that impersonation is not confined to politicians and global celebrities. Ordinary successful creators can also become targets because their social-media archives provide abundant material for AI systems.
Verified expert insights
The clearest expert warning is that deepfakes exploit something humans already rely on: recognition.
Alice Marwick's research highlights why creators are especially vulnerable. Their reputation is directly tied to their identity, meaning an impersonation can potentially damage both audience trust and commercial opportunities.
Australian financial regulators are equally direct. ASIC Chair Sarah Court has warned consumers that polished content, familiar branding and convincing testimonials are no guarantee that an investment opportunity is legitimate.
That advice increasingly applies beyond finance.
A convincing face is no longer proof of authenticity.
Future outlook
The next phase of the creator economy is likely to focus increasingly on proof of identity and proof of consent.
Platforms will continue developing automated detection. Creators will likely demand stronger likeness clauses in contracts. Brands will need clearer rules governing digital replicas, and agencies may eventually treat creator-likeness rights as a standard part of campaign compliance.
The pressure is already visible.
YouTube is developing likeness-detection technology, Meta is expanding impersonation protections, and regulators in countries including Australia and the UK are increasing scrutiny of AI-enabled deception.
But technology alone will not solve the problem.
The most important shift may be cultural. Social-media users are gradually learning that seeing a familiar person on screen is no longer enough to prove that person actually made the content.
That could permanently change how influencer marketing works.
Conclusion
Deepfake influencer advertising has moved beyond an interesting AI experiment. It is becoming a serious challenge to the trust economy that social media creators built.
The technology can produce legitimate creative opportunities, but it can also reproduce a person's identity without permission, manipulate audiences and attach trusted faces to products or scams.
For creators, the message is clear: protecting a digital likeness is becoming part of protecting a business.
For brands, consent and verification are becoming essential.
And for audiences, the safest assumption is increasingly simple: a familiar face online is a reason to investigate, not automatic proof of endorsement.
Haven Trends: Join the conversation
Have you seen an AI-generated celebrity or influencer endorsement that looked genuine at first?
Tell us what you think in the comments, share this article with your social-media community, and subscribe to Haven Trends for more verified coverage of creators, AI, internet culture and the digital trends shaping what comes next.
For more stories on the changing creator economy, social-media platforms and artificial intelligence, keep exploring Haven Trends.
Discover Trending Categories Everyone Is Talking About
💬 Enjoyed the read? Share it forward.
📢 Spread the word, spark a conversation, and Let others stay informed and inspired by what’s shaping the world right now.
💬 Enjoyed the read? Share it forward.
📢 Spread the word, spark a conversation, and Let others stay informed and inspired by what’s shaping the world right now.
Inspire someone else. Share it, start a conservation 👇👇👇👇