Terragon Group "AI Data Annotation Outsourcing" High-Salary Trap: Wage Deductions, Forced Overtime, and Training Fee Scams

Who Gets Targeted

The victims are primarily young job seekers from African countries such as Kenya and Nigeria, mostly aged 18 to 30, with educational backgrounds ranging from high school to incomplete university. Eager to enter the "AI tech industry" and lacking awareness of data annotation work, they are lured by pitches of "high salaries and guaranteed employment after training." Their psychological vulnerabilities include a blind worship of the digital economy and a lack of access to formal employment channels. Combined with high local unemployment rates, they are desperate to seize an opportunity to "make it," often overlooking contract details and even willing to pay "training fees" in advance for a "decent" job.

骗局怎么运作

• Step 1: Post high-paying job openings for "AI Data Annotators" on social media and local recruitment websites, claiming no experience is required and monthly salaries can reach 2 to 3 times the local average, attracting a large number of youth to submit resumes. The recruiters package themselves using titles like "partner of global AI companies" to create a legitimate corporate image.
• Step 2: Arrange online or offline "interviews." The interview process is simple, with almost no technical threshold, focusing heavily on emphasizing that "pre-job training lasting 2 to 4 weeks is required before employment." Training fees range from $200 to $500 per person, with promises of "guaranteed placement and high salaries" after completion.
• Step 3: The training content mostly consists of basic computer operations and simple annotation rules, completely disconnected from the promised "core AI technology." No wages are paid during training, food and lodging are self-funded, and some job seekers are even induced to sign loan agreements to pay the training fees.
• Step 4: After training, job seekers are assigned heavy data annotation work, with daily working hours reaching 10 to 12 hours, far exceeding local legal working hours. Wages are calculated per piece, unit rates for annotations are extremely low, and wages are frequently deducted under the pretext of "failed quality inspections," resulting in actual earnings far below the promised level.
• Step 5: If job seekers raise objections or request to resign, companies withhold their last month's salary under pretexts like "unsettled training fees" or "violation of non-disclosure agreements," and even threaten lawsuits. In some cases, departing workers are required to continue paying "remaining training fees," creating a vicious cycle of "working to pay off debts."

红旗信号（看到这些快跑）

• 🚩 Job postings simultaneously feature buzzwords like "high salary," "no experience required," and "free training included" (or training provided), with salary levels significantly higher than similar local positions.
• 🚩 The interview stage does not assess professional skills, but instead heavily pushes "paid training" or "training loans."
• 🚩 Employment contracts have vague clauses regarding training fees and breach penalties, or oral promises are not put into writing.
• 🚩 The actual work content does not match the job description, turning out to be low-tech, high-intensity repetitive labor.
• 🚩 Delayed wage disbursements, lack of pay slips, or arbitrary salary deductions based on vague reasons like "unqualified performance."

真实案例

• In 2025, according to a report by the OpenDemocracy website, multiple young people in Nairobi, Kenya, applied as annotators for an "AI data company." After paying $300 to $500 in "training fees," they were required to work 12 hours a day, and their actual monthly income was less than one-third of what was promised. Many filed complaints with local labor departments.
• In March 2026, an investigative report on "recruitment-to-training traps" by Xinhua Net disclosed that similar scams were spreading across multiple African countries. Some gangs even colluded with local intermediaries, using the pitch of "striking gold overseas" to lure job seekers to Southeast Asia or the Middle East, where their passports were confiscated and they were subjected to forced labor upon arrival.
• In July 2026, a South African tech and finance website exposed a criminal ring operating under the guise of "AI data outsourcing." After a developer victim was tricked into paying $600 in training fees, they were assigned to annotate inappropriate content for months, and ultimately dismissed under the pretext of "data leakage" without receiving any compensation.
• On June 6, 2024, the People's Court of Chongchuan District, Nantong City, Jiangsu Province, tried an overseas employment fraud case: Three companies used the pretext of "processing overseas labor services with guaranteed high salaries" to defraud a total of 1,152 victims from all over the country, involving 11.65 million yuan. Public prosecutions were initiated against 37 people; in another case disclosed during the same period, a company executive collected a total of 987,000 yuan in overseas employment security deposits from 329 workers nationwide. (Source: [https://www.news.cn/legal/20260703/5c53b78932864289b7503acb6330b1b3/c.html](https://www.news.cn/legal/20260703/5c53b78932864289b7503acb6330b1b3/c.html))
• In May 2023, in an overseas employment fraud case prosecuted by the People's Procuratorate of Fangshan District, Beijing, over 260 victims were defrauded of nearly 10 million yuan by criminals. Court trials revealed that between 2021 and 2024, the intermediary companies involved used similar methods to defraud dozens of victims out of a total of over 1.02 million yuan, with individual job seekers losing anywhere from several thousand to tens of thousands of yuan. (Source: [https://www.news.cn/legal/20260703/5c53b78932864289b7503acb6330b1b3/c.html](https://www.news.cn/legal/20260703/5c53b78932864289b7503acb6330b1b3/c.html))

Official Stance

• On July 3, 2026, Xinhua Net published "Xinhua Investigation: Beware of Overseas Job-Hunting 'Gold-Digging Traps'," reminding job seekers to watch out for overseas employment scams under the guise of "high salaries" and "training," and pointing out that emerging industries like data annotation have become heavy disaster areas for "recruitment-to-training" schemes.
• On July 17, 2026, the National College Student Employment Service Platform (tongji.ncss.cn) issued a special advisory on "Beware of Overseas Job-Hunting Gold-Digging Traps," urging young job seekers to enhance risk awareness and avoid prepaying any fees to employers.
• On March 31, 2026, Xinhua Net published "High-Salary Internal Recommendations Hide Traps: 'Recruitment-to-Training' Scams Target Job Seekers," pointing out that some "data annotation" companies collect exorbitant fees under the guise of "internal referrals" while actually engaging in labor exploitation.

How to Protect Yourself

• ✅ Verify company qualifications before job hunting, check their business scope and establishment date on government commercial registration websites, and be wary of "shell companies" that have been established for less than a year and lack physical office locations.
• ✅ Refuse any form of "pay-to-work" or upfront fees; legitimate enterprises never charge job seekers training fees, deposits, or uniform fees.
• ✅ When signing employment contracts, carefully review clauses regarding training, salary, working hours, and resignation. Request written clarification for vague wording and keep the original contract.
• ✅ In the event of wage deductions or forced overtime, immediately file a complaint with local labor supervision authorities and collect evidence such as pay slips, chat records, and attendance logs.