Amazon says it is all social media’s fault for letting faux overview schemes thrive • The Register

Amazon, which has struggled to curb counterfeiting and fraud on its e-commerce platform for years, has blamed social media companies for undermining its integrity efforts.

In a post related to its Project Zero anti-counterfeiting initiative launched in 2019, the computer rental and gift business celebrated its focus on customers and reprimanded unidentified social media companies for failing to crack down on vendors of fake product reviews.

Coincidentally, Amazon’s lobbying is said to have helped thwart the Inform Consumers Act, which proposed a counterfeiting law that would have required Amazon to authenticate third-party merchants on its platform. The company claims the legislation is favorable to brick and mortar stores.

Either way, Amazon said that while it has developed and used technology “over many years” to prevent fake reviews from ever being seen, it recognizes that its approach has limitations, which is why it has hired customers and sellers to report fake reviews – without paying for their time or work.

But even those efforts have proven inadequate, the company said, as review manipulators switched to social media to enable what Facebook would call “coordinated inauthentic behavior.”

“Because of our continued improvements in detecting fake reviews and connections between buying and selling accounts by malicious actors, we’ve seen an increasing trend of malicious actors trying to solicit fake reviews outside of Amazon, particularly through social media. Services, ”said the company.

“Some use social media services themselves; in other cases, they hire a third party to perform this activity on their behalf. However, malicious actors regularly attempt to conduct this transaction outside of Amazon to hide our ability to detect their activity.” and the relationship between the multiple accounts that commit or benefit from this abuse. “

Amazon says it identified more than three times as many groups associated with fake reviews (1,000+) in the first three months of 2021 as it did in the same period of 2020 (300+) and the ecommerce Business praised social media companies to respond faster to abuse reports. Average response time to closing groups alleged violating rules dropped from 45 days last year to just five days this year. That matters something.

Nonetheless, Amazon wants social media companies to take unsolicited action and monitor the platform’s activities.

“While we know some social media companies are reacting much faster, it is imperative for social media companies to invest appropriately in proactive controls to detect and enforce fake reviews before we report the problem to them Tackle problem on a large scale, “it said.

Amazon hasn’t identified which social media companies it has in mind, but security researchers have observed that Facebook and its Instagram subsidiary play an important role in creating and disseminating fake reviews, as they do with fake news .

In a pre-printed paper [PDF] Rajvardhan Oak, a graduate student at UC Berkeley who later joined Microsoft, wrote, “Our research shows that Facebook played an important role in this, which was distributed through ArXiv earlier this year:“ The Fault in the Stars: Understanding the Underground Market of Amazon Reviews “. the continued operation of the underground criticism market. “

Facebook does this with the help of ads. “Targeted advertising makes it easier to identify users who might be complicit in writing incentive-based reviews and gradually draws them deeper into the underground market,” explains Oaks paper.

This approach, described in a research paper from 2019 [PDF] as a “crowdturfing attack” can also be carried out using neural language models such as GPT-2 NLM.

“Because this method has an economic cost, it is usually limited to large-scale attacks,” explains the machine learning paper. “Automated crowdturfing, which uses machine learning algorithms to generate fake reviews, is a cheaper and more efficient way to attack online review systems.”

With Facebook reporting that it removed 1.3 billion fake accounts in the first quarter of 2021, up from 1.7 billion in the first quarter of 2020, some progress may be made. However, Facebook does not disclose how long these accounts were active or how they were used during that time. Judging by Amazon’s own seller forums, fake reviews are still a problem.

As a result, Amazon has asked for “coordinated help from consumer protection agencies around the world”.

Sonos sings

Separately, in Washington, DC, on Wednesday before a Senate Justice Antitrust Subcommittee, executives from Amazon [PDF] and google [PDF] took turns defending their home-oriented smart speaker technology (Alexa and Google Assistant) and observing how highly competitive the market is (despite pending antitrust complaints against each company).

A Sonos manager [PDF] meanwhile criticized the anticompetitive dominance of Big-Tech and pointed out that Google Assistant cannot run on its speakers at the same time as other assistant software because Google would not allow this.

There were also academics who urged to do something without being too sure about certain remedies.

“Today the platform companies appear to many as an imperial power,” noted Matt Crawford, Research Fellow at the Institute For Advanced Studies In Culture, in a prepared statement [PDF]. “The fundamental question of ‘Who rules?’ is pressed again on this body. “

This is a provocative question in a room full of rule makers, some of whom differ on who is legitimately running the United States. But it’s worth asking. ®