China's Anti-Fraud Center Flags Dove's Apology as AI-Generated
After Dove apologized for a widely condemned Qixi marketing line, China's National Anti-Fraud Center said its own AI test flagged the apology as machine-generated — turning a branding stumble into a referendum on corporate crisis PR in the age of large language models.
Dove's apology for a tone-deaf Qixi campaign has become a different kind of case study: a public experiment in how Chinese consumers judge the difference between a company's remorse and a chat bot's. After the brand's statement landed, China's National Anti-Fraud Center said its tool had flagged the text as AI-generated, and the finding raced up Weibo's trending list under the slogan "the anti-fraud center detected that Dove's statement is AI."
The saga began on Aug. 23, when a holiday-themed short video on Dove's official flagship store told a would-be customer that the way to get the Qixi gift box she wanted was not to ask her boyfriend but to visit an "ambiguity partner" — a romantic-but-not-defined acquaintance — and then drive two hours, implying the flirtation would bring the chocolate. Qixi, celebrated in late August this year, is China's Valentine's Day, and the video's premise — pitting a real boyfriend against a side flirtation for a box of chocolate — struck critics as celebrating precisely the kind of disloyalty the holiday is not about. Chinese media commentary piled on, arguing that brand "meme-chasing" had collided with basic public decency.
Dove's apology, published to its official account on Aug. 25, said it took full responsibility and that the video "was created and published by a third-party store operations team that failed to follow the brand's content review process." Then came the twist: the anti-fraud center, a government body better known for warning citizens against telecom scams than for brand arbitration, fed the statement through its AI-detection tool and said it bore the fingerprints of machine text. Commenters read the result as proof of insincerity: the company had apologized, but an algorithm had done the apologizing.
The episode fits a longer pattern on Chinese social media, where corporate "翻车" — literally "rolling over," slang for a public blunder — is treated as a test of authenticity. Brands that rush out a smooth, cover-all-bases statement increasingly get rewarded with a new accusation: that the words were produced by a model and hollowed of any human accountability. The practical lesson being drawn this week, in media commentary and comments alike, is not limited to Dove. As generative AI becomes ubiquitous inside marketing departments, the question Chinese consumers are now equipped to ask of any crisis response is whether a human being actually read it before the company hit send.