Tai Chi Ads Illustrate AI-Enabled Scam Tactics
Creative Bloq reported on July 26 that ubiquitous AI-generated tai chi walking ads show scammers can use visibly artificial imagery to mislead viewers, not just photorealistic deepfakes. The Tai Chi Notebook described ads featuring a muscular older Asian man and claims that tai chi walking can produce a six-pack, calling them obviously AI-created and fraudulent; the retrieved coverage does not identify their operators.
Creative Bloq reported on July 26 that AI-generated "tai chi walking" advertisements demonstrate a less discussed fraud risk: deceptive ads need not look fully realistic to attract attention and mislead viewers. The article contrasts concern over photorealistic AI images and video with the continued spread of ads whose synthetic visual qualities are readily apparent.
The Tai Chi Notebook, writing in April, described a recurring version of the ads as featuring a well-muscled older Asian man and promoting the claim that tai chi walking can deliver a six-pack. The blog characterized the adverts as obviously AI-created and fraudulent, while noting that their apparent artificiality had made their distribution seem counterintuitive.
Deception beyond photorealism
Creative Bloq's reporting frames the ads as evidence that visual realism is not the sole mechanism behind successful AI-assisted deception. Synthetic-looking material can still be used in misleading advertising, particularly when paired with extraordinary health or fitness claims.
The retrieved coverage does not identify the people or organizations operating the ads, establish a verified financial total, or document the advertising platforms involved. It also does not provide independent evidence about conversion rates or victim losses.
For ML and trust-and-safety practitioners, the case illustrates a broader moderation pattern: detection systems focused narrowly on whether an image is a convincing deepfake can miss harmful content whose risk lies in the accompanying claim, targeting, and distribution behavior. In comparable scam campaigns, multimodal review of creative assets, landing pages, advertiser provenance, and repeated claim templates can be more informative than a photorealism test alone.
Moderation implications
The case also shows why clearly visible generation artifacts are not, by themselves, reliable safety cues. Labeling generated media addresses provenance, but it does not independently establish whether an accompanying health, financial, or product claim is deceptive.
The available reporting therefore supports a narrow conclusion. AI image generation can support misleading ad creative even when its output remains visibly synthetic; the sources do not establish the scale or operators of this specific ad network.
Key Points
- 1Creative Bloq reports that visibly synthetic AI imagery can support scam advertising, challenging the assumption that fraud requires photorealistic deepfakes.
- 2The Tai Chi Notebook documented recurring fitness ads that pair AI-generated characters with implausible claims about tai chi and six-pack results.
- 3The case shows that image realism alone is not a sufficient indicator of whether advertising content is deceptive.
Scoring Rationale
The story identifies a practical trust-and-safety issue: visibly artificial generative media can still be effective scam creative. Its direct practitioner impact is moderate because the retrieved reporting does not establish the campaign's operators, platform reach, losses, or a new technical exploit.
Sources
Public references used for this report.
Practice with real Ad Tech data
90 SQL & Python problems · 15 industry datasets
250 free problems · No credit card
See all Ad Tech problems
