AI Intelligent Computing Center GPU Computing Power Hosting Dividend Scheme: Absorbing public funds under the guise of purchasing computing power server hosting for dividends, with over 100 million RMB deposited by 110,000 members before withdrawals were frozen

Who Gets Targeted

Primarily investors aged 30 to 55 with some savings but a lack of fundamental technical judgment, including small business owners, retired technical personnel, veteran crypto retail investors, and white-collar workers attracted by the 'AI computing power dividend' narrative. They generally share three psychological vulnerabilities: first, anxiety about the AI large-scale model wave and the fear of missing out on the next wealth windfall; second, the alternative psychological drive of 'if you can't buy the stocks, buy the computing power' after seeing stock prices of computing giants like NVIDIA surge; third, a false sense of security derived from the tangible, visible packaging of 'physical server hosting,' believing it differs from purely virtual coin schemes and thereby lowering their guard. Instructors in communities showcase forged server room photos and dividend transaction streams, further reinforcing herd behavior.

骗局怎么运作

• Step 1: Packaging computing power assets. The platform claims to have built intelligent computing centers in Inner Mongolia, Guizhou, or overseas, equipped with thousands of A100 or H100 graphics cards. Investors can purchase 'computing power server shares' in packages, with unit prices ranging from several thousand to hundreds of thousands of RMB. After purchase, the platform assumes full custody and operational management, requiring no self-management of hardware by investors.
• Step 2: Committing to fixed dividend returns. Promotional materials indicate a monthly yield of 8% to 15%, with dividends distributed daily or weekly. Early investors do indeed receive small rebates; this capital actually originates from the principal of later entrants, aiming to create the illusion of real dividends and incentivize investors to purchase larger packages.
• Step 3: Community fission recruitment. The platform establishes a referral reward mechanism where investors inviting new users to purchase computing power shares earn a commission of 10% to 20% of the downline's invested amount, alongside team-tier rewards, thereby forming a multi-level promotion structure. Instructors regularly post screenshots of tens of thousands in daily dividends and withdrawal arrival records in communities to attract onlookers to deposit funds.
• Step 4: Manufacturing a sense of urgency regarding computing power shortages. Pitches repeatedly emphasize that 'NVIDIA graphics card production capacity is limited, global computing power supply falls short of demand, and if you don't lock in computing power shares now, you will only be able to rent at high prices in the future,' while regularly publishing forged computing power order data and partner enterprise lists to induce anxiety that 'if you don't buy now, you'll miss out,' driving rapid decision-making.
• Step 5: Position locking and absconding. When the rate of new capital inflow falls below dividend payout expenditures, the platform suddenly issues a notice suspending withdrawals due to 'policy compliance reviews' or 'computing power scheduling maintenance,' subsequently closing its official website and communities. Operators transfer funds to overseas wallet addresses, and the ownership of the so-called computing power shares held by investors becomes untraceable, making it impossible to seek legal recourse or locate physical assets.

红旗信号（看到这些快跑）

• 🚩 Claiming monthly returns of over 8% with fixed payoffs, yet failing to provide any independently verifiable computing power output data and customer settlement records, dodging reasonable requests to visit physical server rooms or refusing on security grounds
• 🚩 Requiring investors to transfer funds into personal bank accounts or non-platform corporate accounts rather than legitimate third-party escrow accounts, with contract entities being recently registered shell companies rather than long-standing technology enterprises
• 🚩 Recommendation reward tiers exceeding two levels, with commission ratios significantly higher than industry norms, clearly relying on new recruitment funds to cover dividend payouts rather than computing power operational revenue
• 🚩 Claiming partnerships with NVIDIA or prominent cloud vendors without being able to provide authorization documents; platform-displayed server room photos, upon reverse searching, turn out to be public stock photo assets or highly similar to official photos of real data center websites
• 🚩 Platform operating entities frequently changing company names or registered addresses, with official website domain registration times extremely short despite claims of multi-year operations, and community administrators knowing nothing about technical details while repeatedly emphasizing yield figures

真实案例

• According to a 2024 report by Panxun Network, 'Lianhe Intelligent Computing Center' grew to 110,000 members under the guise of computing power hosting dividends. After operators amassed over 100 million RMB, withdrawals were halted and the scheme collapsed as operators absconded. The report noted member investment amounts ranging from thousands to hundreds of thousands of RMB, with victims currently seeking rights protection
• According to a 2024 report by Piyao Network, the 'Qiyuan AI' project attracted tens of thousands of members to deposit funds under the names of AI avatars and computing power hosting, but investors were banned upon attempting to withdraw. The report pointed out that project organizers claimed monthly incomes exceeding 10,000 RMB without any genuine computing power operational support
• According to a 2024 early-warning report by Shima Home, 'Aqi AI' claimed monthly returns exceeding 30%. Verification showed it was a reskinned derivative of the collapsed ARK project, with the report noting that the project used AI computing power investment packaging to execute a secondary harvest

Official Stance

• In 2021, six departments including the China Banking and Insurance Regulatory Commission jointly issued the 'Risk Warning on Preventing Illegal Fund-Raising under the Guise of Virtual Currency and Computing Power,' explicitly stating that acts promising high returns under the guise of computing power investments and virtual currencies involve suspected illegal fund-raising
• In 2022, the National Internet Finance Security Technology Committee issued a warning pointing out that certain platforms absorbed public funds using 'AI computing power hosting' as a gimmick, which in essence were Ponzi schemes, reminding the public to be vigilant against fixed high-yield promises
• In 2024, the Shenzhen Public Security Bureau formally initiated a criminal investigation into the 'Zhongzheng Ming Quantitative AI Wealth Management Scheme,' with estimated involved funds exceeding 300 million RMB. The case absorbed public funds under the guise of AI quantitative wealth management before executing a comprehensive position lock

How to Protect Yourself

• ✅ Require platforms to provide third-party auditable computing power operational qualifications and original technical cooperation agreements, verifying whether the claimed chip manufacturers or cloud service providers have genuine authorization relationships by directly calling manufacturer customer service
• ✅ Refuse any requests to transfer funds into personal accounts or non-corporate accounts; legitimate computing power hosting services should settle through corporate bank accounts or licensed third-party payment channels and issue standard invoices
• ✅ Before investing, check the industrial and commercial registration information of the platform operating entity on official public registration information websites, focusing on registration date, paid-in registered capital status, and whether the business scope includes computing power services, while cross-checking the shareholder list and historical change records
• ✅ Maintain vigilance against projects with monthly yields exceeding 5%, and estimate true computing power yield margins by comparing market prices for equivalent GPU rentals from AWS, Alibaba Cloud, etc.; if advertised yields vastly exceed market levels, the scheme can basically be judged as a Ponzi scheme