A study revealed the existence of a vast market that buys false reviews to boost sales of products on Amazon, produced by unknown companies in China, generally.

Researchers from the University of Southern California,  and the University of California at Los Angeles uncovered the salesmen’s plan to trick Amazon customers with mediocre products, The Economist reported on Sept. 3.

The report said it found more than 20 Facebook groups dedicated to selling false recommendations, with an average of 16,000 members. In more than 560 posts each day, sellers offered to pay an average of $6 for each misleading review.

Quarantines caused by the CCP (Chinese Communist Party) Virus have boosted sales disproportionately, but those misleading ads hurt buyers with dubious products and legitimate sellers, some of whom have pulled out of Amazon.

The Financial Times also investigated the status of misleading recommendations in the United Kingdom and found irregularities in 9 of the top 10 contributors of these ads in the country to Amazon.

Even the top UK Amazon reviewer, Justin Fryer, reported on products worth more than $16,000 in August alone, scoring five stars every four hours on average.

Chinese companies often offer samples of their products in return for fake reviews, and Justin Fryer has earned over $22,000 since June by selling products he reviewed on Amazon on eBay.

Fryer’s fraudulent schemes had already been discovered by one of the Amazon users but his complaints were ignored, so he wrote directly to Jeff Bezos in early August.

The online review analysis group Fakespot estimated that the scam peaked in May when 58 percent of the products on Amazon.co.uk had apparently false reviews.

“The scale of this fraud is staggering,” said Saoud Khalifah, CEO of Fakespot. “And Amazon UK has a much higher percentage of fake reviews than the other platforms,” according to the Financial Times on Sept. 4.

The business appears to be so profitable that numerous robots are deployed to issue false recommendations. One creator of such a robot said he had processed more than 16,000 five-star reviews in the last year, according to the Financial Times.