Trustpilot AI Fake Review Reverse Extortion - The Grey Industry Chain of AI Generating Bulk Fake Positive Reviews and Charging to Delete Negative Reviews

Who Gets Targeted

Victims fall into two main categories: First, regular consumers who rely on online reviews to make purchasing decisions, as they are misled by AI-generated fake positive reviews into buying inferior goods or services. Second, extorted small and medium-sized business owners, particularly cross-border sellers, who worry that negative reviews will harm traffic and conversion rates, often feeling forced to pay thousands to hundreds of thousands of yuan upon receiving threats of paid negative review deletion. These victims share common vulnerabilities: trust in platform review systems, fear of direct business losses caused by negative reviews, and a lack of awareness regarding the AI water army industry chain, making them easily manipulated by scam syndicates amidst anxiety and urgency.

骗局怎么运作

• Phase 1: Automated account farming and content factories. Scam syndicates use AI tools to batch-create fake consumer accounts, automatically generating large volumes of coherent, detailed positive and negative reviews via natural language generation models (such as GPT-class models) at a cost of only 5 to 15 dollars per review. They combine this with proxy IPs and anti-detect browsers to simulate real user behavior and evade platform risk control.
• Phase 2: Targeted deployment of negative reviews. Targeting the enterprise, the syndicate intensively publishes a large number of 1-star negative reviews within a short period. The content mostly consists of generalized accusations such as 'poor quality,' 'unresponsive customer service,' or 'slow shipping,' and some negative reviews are accompanied by forged order screenshots or chat logs to enhance credibility.
• Phase 3: Reverse extortion contact. The syndicate contacts the enterprise via email, WhatsApp, or third-party platforms, claiming to be able to 'clear negative reviews' or 'optimize ratings,' with quotes ranging from thousands to hundreds of thousands of yuan, while threatening that refusing to pay will result in additional negative reviews or reporting the enterprise to regulators for 'genuine issues.'
• Phase 4: Cyclical arbitrage and secondary harvesting. If the enterprise pays, the syndicate temporarily deletes some negative reviews but subsequently demands additional payments under pretexts such as 'requiring protection fees to maintain service' or 'residual data remaining,' and may even sell the enterprise information to black-market databases, triggering further scam harassment.
• Phase 5: Whitewashing and scaling. The syndicate uses AI to rapidly replicate the entire process, operating in batches targeting cross-border e-commerce, overseas independent websites, and other industries relying on Trustpilot and Google ratings, while collecting payments in virtual currency to evade regulatory scrutiny.

红旗信号（看到这些快跑）

• 🚩 A sudden surge of 5-star or 1-star reviews within a short period that share identical content and templated sentence structures, with the reviewers' avatars/nicknames being random and showing no genuine purchasing behavior
• 🚩 Mentioned 'order numbers' or 'logistics tracking numbers' in negative reviews that cannot be verified in the platform system or do not match real purchase records
• 🚩 Individuals proactively contacting you through unofficial channels (email/direct message) promising to 'delete negative reviews' and guiding you toward private transfers or cryptocurrency payments
• 🚩 Service providers claiming '100 percent negative review removal' or 'results within 24 hours' but refusing to sign a formal contract or provide a genuine office address
• 🚩 Review-publishing IPs being concentrated, times being unusually dense (such as between 2 AM and 5 AM), or review content containing AI-common vocabulary such as 'in summary' or 'furthermore'

真实案例

• In March 2026, the Italian Antitrust Authority (AGCM) fined Trustpilot and its subsidiary 4 million euros, ruling that it failed to adequately verify the authenticity of reviews and that its service operation methods were misleading, directly exposing platform moderation loopholes. (Source: [https://mp.weixin.qq.com/s?__biz=MzI1MTM2ODk0OA%3D%3D&chksm=e8c268c6257a25f01a51600b295e31d4f043d8eb81af54d7e2dd5b52723fe78c72c1c40f0eb6&idx=2&mid=2247558416&scene=27&sn=2158eb06309575df1adcd8d6393c1972](https://mp.weixin.qq.com/s?__biz=MzI1MTM2ODk0OA%3D%3D&chksm=e8c268c6257a25f01a51600b295e31d4f043d8eb81af54d7e2dd5b52723fe78c72c1c40f0eb6&idx=2&mid=2247558416&scene=27&sn=2158eb06309575df1adcd8d6393c1972))
• Shanghai police recently cracked down on multiple cases of 'public opinion extortion,' where criminal syndicates hired internet water armies to publish over a thousand fake negative reviews to smear competitors, and used AI tools to fabricate false articles concerning enterprises, demanding 2.3 million yuan in 'consultation and post-deletion fees.' Multiple suspects have been subjected to criminal compulsory measures (according to official announcements).
• Huxiu reported in May 2026 that CCTV exposed e-commerce platform review sections flooding with AI-generated fake buyer shows, where merchants used AI to generate exquisite images and reviews to boost ratings, while professional negative reviewers used AI to forge damage evidence to extort refunds from sellers, with some sellers having been defrauded of thousands of yuan. (Source: [https://mp.weixin.qq.com/s?__biz=MzI1MTM2ODk0OA%3D%3D&chksm=e8c268c6257a25f01a51600b295e31d4f043d8eb81af54d7e2dd5b52723fe78c72c1c40f0eb6&idx=2&mid=2247558416&scene=27&sn=2158eb06309575df1adcd8d6393c1972](https://mp.weixin.qq.com/s?__biz=MzI1MTM2ODk0OA%3D%3D&chksm=e8c268c6257a25f01a51600b295e31d4f043d8eb81af54d7e2dd5b52723fe78c72c1c40f0eb6&idx=2&mid=2247558416&scene=27&sn=2158eb06309575df1adcd8d6393c1972))

Official Stance

• In March 2026, the Italian Antitrust Regulatory Authority (AGCM) fined Trustpilot 4 million euros, warning platforms to earnestly verify review authenticity.
• Shanghai police (2026) issued consecutive alerts regarding enterprise-related online 'black mouth' cases, explicitly cracking down on the use of AI to fabricate false information and engage in extortion under the guise of deleting posts, and reminding enterprises to report such extortion to the police immediately.
• In August 2026, Amazon and Google jointly upgraded their AI fake review detection systems, officially stating expectations to intercept over 200 million fake reviews and calling on sellers to handle negative reviews through legitimate channels.

How to Protect Yourself

• ✅ When facing a sudden outbreak of negative reviews, immediately save all screenshots, emails, and chat logs, submit evidence through the platform's official appeal channels, and avoid private payments.
• ✅ For service providers who proactively contact you to 'delete negative reviews,' request their company registration information, office address, and successful case studies, check whether they have official ICP filings or overseas compliance qualifications, and reject any cryptocurrency transactions.
• ✅ Establish an enterprise review monitoring system, use official APIs or third-party compliant tools to track rating changes in real time, distinguish normal negative reviews from malicious attacks, and report suspicious bulk reviews to platforms and police in a timely manner.
• ✅ Cross-border sellers should explicitly prohibit service providers in contracts from adopting non-compliant post-deletion methods, regularly train customer service to identify AI phrasing (such as templated structures in abnormal praise or criticism), and back up all operational data to prevent extortion.