EduLab AI: Pirated Question Bank Disguised as Personalized Learning System Scam

Who Gets Targeted

Victims are mainly divided into two categories: The first category consists of K12 student parents, who generally harbor anxiety about entering higher education and hold extremely high expectations for "AI precise score boosting." They are easily swayed by sales pitches such as "elite school internal question banks" and "big data analysis of weak points." Lacking technical discernment, they assume paying money grants exclusive resources. The second category consists of academic administrators in public or private schools. Facing pressure from digital transformation performance assessments and hoping to rapidly introduce "advanced AI systems" as teaching highlights, they neglect due diligence regarding content copyright and algorithmic authenticity. Fraudsters exploit the psychology of "authoritative endorsement"—falsely claiming cooperation with elite schools or research institutes, combined with forged contracts and success cases—causing school authorities to lower their guard during collective decision-making, ultimately resulting in both budget losses and damage to students' learning rights.

骗局怎么运作

• Step 1: Product packaging. The fraud ring registers shell companies like "EduLab AI," creates a polished official website, displays false concepts such as "adaptive learning engines" and "brain science algorithms," and claims cooperation with multiple key middle schools. In reality, the backend contains only a simple keyword-matching recommendation script without any genuine learning models. They hire outsourcing teams to produce product demonstration videos, using preset scripts to fake student grade improvement curves, and embed unverified "student counts" and "score-boosting rates" into promotional materials to provide dual visual and data endorsements for subsequent sales to schools.
• Step 2: Targeted customer acquisition. Through educational exhibitions, agent networks, and online information stream ads, they precisely reach middle/primary school principals or academic affairs directors. Sales pitches emphasize the "National Education Digital Strategy" and "falling behind if you don't upgrade," promising free trial accounts and "expert on-campus guidance" services. In reality, the trial period only opens simple random pushes of a small amount of pirated question banks, masking the system's true capabilities. Simultaneously, they leverage parent chat groups to spread the bandwagon effect that "classmates are all using it," creating false urgency and prompting schools to sign annual subscription contracts without deep testing.
• Step 3: Padding with pirated question banks. The backend database pirated and pieced together mock exams, cram school book scanned copies, and circulating past exams from various regions without any copyright authorization. The system merely tags data coarsely by "grade + subject + difficulty." The so-called "personalized push" randomly selects questions from the question bank and uses preset scripts to generate fake reports claiming "this practice targeted your weak knowledge points." In reality, it collects no student response process data and provides no intelligent diagnosis. All reports are abstracted into uninformative charts like "accuracy rate increased by X%," leading parents and teachers to mistakenly believe the algorithm is optimizing the learning path, whereas it merely counts click frequencies.
• Step 4: Billing and renewal traps. Using tiered pricing such as "Basic version free, Premium version 29,800 yuan/year," the premium version promises "elite school teacher AI customized schedules and exclusive learning analytics." Contracts specify a "technical support fee" but omit specific content update frequencies. Once schools or parents request refunds, customer service refuses on the grounds that "technical services have been provided" or sets "auto-renewal after trial period" clauses, using standard contract traps to make it difficult for users to protect their rights. In actual operations, the platform updates only a small amount of public resources monthly, the overwhelming majority of pirated question banks come from one-time scraping with no subsequent investment, and the company further develops downlines through a "regional exclusive agency" model, forming a pyramid distribution where agents prepay franchise fees and bear customer acquisition costs themselves, intensifying tiered harvesting.
• Step 5: Absconding and shifting blame. After operating for several months to a year, once schools discover poor system quality and parent complaints increase, the company deregisters the original entity and re-registers new brands like "Xuehui AI" or "Zhixue Yun" to continue the scam. Outwardly, they claim "server upgrades" or "algorithmic iteration"; in reality, they transfer assets before legal risks are exposed, leaving behind unfulfillable service contracts and pirated content. Some companies also shift blame to "technical partners," claiming recommendation algorithms were developed by third parties and they only handle integration, thereby evading direct infringement liability—even though all stages were solely orchestrated by them.

红旗信号（看到这些快跑）

• 🚩 Claims "exclusive" or "top-secret" question banks but cannot provide copyright proof, and identical content can be found for free on the public internet.
• 🚩 Contract clauses overly emphasize "technical support fees" while remaining vague on content copyright, data privacy, and refund rules.
• 🚩 Provides only demo accounts and refuses to let the school's technical team conduct source code or algorithmic audits, with so-called "personalization" entirely based on sales talk descriptions.
• 🚩 Uses promissory vocabulary such as "score-boosting rate" and "prediction hit rate" in promotions, fabricating endorsement partnerships with well-known universities or research institutions.
• 🚩 Demands that schools sign multi-year contracts and prepay the full amount all at once, rather than paying per semester or based on results.
• 🚩 Customer service and technical support contacts are unstable, agents change frequently, and the company's registered address does not match its operating address.

真实案例

• In June 2026, the official website of the Ministry of Education issued a Gaokao warning, reporting that "unlawfully operating individuals impersonated proposition committee teachers to sell 'top-secret past exams' and 'AI prediction papers,' with involved amounts exceeding 10 million yuan." The Smart Legal Services Platform simultaneously published the warning, reminding candidates and parents to beware of false promotions such as "AI exam predictions." After paying 20,000 yuan for so-called prediction services, a student's parent in a certain province discovered all questions were public old questions. (Source: [http://legalinfo.moj.gov.cn/pub/sfbzhfx/zhfxfzzx/fzzxyw/202606/t20260602_535928.html](http://legalinfo.moj.gov.cn/pub/sfbzhfx/zhfxfzzx/fzzxyw/202606/t20260602_535928.html))
• On July 6, 2026, the Cyberspace Administration of China announced the progress of the first phase of the "Qinglang - Rectifying AI Application Chaos" special campaign: over 14,000 irregular websites, applications, and smart agents of AI products were handled, more than 6 million pieces of illegal or irregular information were cleaned up, over 26,000 accounts were handled, 1,300+ illegal AI commodities were removed from shelves, and 9 illegal open-source datasets were addressed. (Source: [https://www.cac.gov.cn/2026-07/06/c_1785081384384987.htm](https://www.cac.gov.cn/2026-07/06/c_1785081384384987.htm))
• In May 2026, media exposed an "AI intelligent paper-composition system" sold under the guise of school procurement to education bureaus across multiple regions. The backend actually loaded only public mock exams from 2018–2022 and lacked any personalized analysis functions. Multiple schools jointly sued requesting refunds. The court ruled the company committed fraud, ordering the return of all fees and compensation for losses, but company assets had already been transferred, making enforcement difficult.

Official Stance

• On June 2, 2026, the Ministry of Education issued the "2026 Gaokao Early Warning Information," explicitly pointing out the need to guard against false promotions like "elite teacher exam predictions" and "AI exam predictions," reminding candidates and parents to take exams honestly and beware of scams.
• On July 6, 2026, the Cyberspace Administration of China released the achievements of the first phase of the "Qinglang - Rectifying AI Application Chaos" special campaign, focusing on rectifying AI data poisoning and missing generated content identifiers, setting up a reporting section, and requiring all AI applications to fulfill filing obligations.
• On June 8, 2026, the Ministry of Justice's Smart Legal Services Platform forwarded the Ministry of Education warning and invited scholars to interpret it, reminding educational institutions everywhere to verify filing and copyright qualifications when procuring AI products.

How to Protect Yourself

• ✅ Before schools procure AI products, they should require suppliers to provide software copyright certificates, content copyright authorization letters, and large model filing numbers, verifying authenticity through the official website of the Cyberspace Administration of China.
• ✅ Ensure a trial period of no less than 1 full semester, require suppliers to provide algorithmic documentation manuals, and allow the school's technical team or third-party institutions to conduct black-box testing to verify whether "adaptability" is genuine.
• ✅ Explicitly stipulate content update frequencies, copyright warranty clauses, and refund mechanisms in contracts, strictly prohibiting vague expressions of "technical support fees," and agree upon liquidated damages.
• ✅ Before parents purchase any "AI score-boosting" service, check the blacklists of the Ministry of Education and the State Administration for Market Regulation, stay alert to "guaranteed score" and "exam prediction" promises, keep payment vouchers and communication screenshots, and immediately call 12315 to complain or report to the police upon discovering anomalies.