At a time when AIGC technology has deeply penetrated content creation, how to safeguard the authenticity of evaluation systems has become the core battlefield for content platforms.

On March 23, Dianping officially released the "2025 Evaluation Transparency Report." The report shows that in the past year, the App received nearly 450 million user reviews, covering almost 9.029 million businesses worldwide. Faced with the growing prevalence of non-original content, Dianping has disclosed governance data on AIGC violations for the first time.

Beating AI with AI: 11.61 million "machine-written reviews" intercepted

To combat mass-produced "cyber soldiers," Dianping introduced an AI-powered review mechanism in 2025. This technological upgrade achieved a high-dimensional scan of review authenticity:

Layered Disposal: According to the report, the platform dealt with a total of 11.61 million AIGC violation reviews last year.

Total Achievements: Combined with the "double insurance" of AI-powered review and manual audit, the platform handled more than 25.57 million types of violation reviews throughout the year.

Cracking down on "Gifts for Reviews": Merchant Numbers Who Seek Positive Reviews Drop by 40%

In addition to combating machine-generated content, Dianping also upgraded its rules to address the common "positive reviews for gifts" phenomenon offline. By optimizing the star rating algorithm, the platform further reduced the impact of incentivized reviews on the real scores of merchants. Data shows that under the intensive special governance, the number of merchants frequently using improper means to promote reviews dropped by 42% compared to the same period last year.

Guarding the "Authenticity" Moat in the AI Era

With the popularization of large models, low-cost "pseudo-real" content is challenging the foundation of internet trust. Dianping disclosed this governance data to send a message to the outside world: in the era of algorithm recommendations, authentic consumer decision-making references remain the top priority for the platform.

From establishing "four lines of defense" to fully integrating AI-powered review mechanisms, Dianping is reshaping the review ecosystem through technical means. When AI can detect AI's disguise, this battle for authenticity has truly entered the second half of the symmetric technology phase.