Amazon says it's all social media's fault for letting fake review schemes thrive

Won't someone else save us from our policies and our decision to forego editorial control?


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

In a post associated with its anti-counterfeiting Project Zero initiative, launched in 2019, the computer rental and gifts biz celebrated its focus on customers while chiding unidentified social media firms for failing to take action against peddlers of fake product reviews.

Coincidentally, Amazon lobbying is said to have helped nix the Inform Consumers Act, which proposed anti-counterfeiting legislation that would have required Amazon to authenticate third-party merchants on its platform. The company claims the legislation favored brick-and-mortar businesses.

Anyway, Amazon said that while "over many years," it has developed and deployed technology to prevent fake reviews from ever being seen, it recognizes that its approach has limits, which is why it has deputized customers and sellers to report fake reviews - without paying for their time or labor.

But even these efforts have proven insufficient, the company said, because review manipulators have shifted to social media to facilitate what Facebook would call "coordinated inauthentic behavior."

"Due to our continued improvements in detection of fake reviews and connections between bad-actor buying and selling accounts, we have seen an increasing trend of bad actors attempting to solicit fake reviews outside Amazon, particularly via social media services," the company said.

"Some use social media services on their own; in other cases, they hire a third-party service provider to perpetrate this activity on their behalf. However, bad actors regularly try to take this transaction outside Amazon to obscure our ability to detect their activity and the relationship between the multiple accounts committing or benefiting from this abuse."

Amazon says it identified more than three times as many groups associated with fake reviews in the first three months of 2021 (1,000+) than it did in the same period of 2020 (300+) and the e-commerce biz lauded social media companies for responding to abuse reports more quickly. The median response time to shut down groups accused of rules violations dropped from 45 days last year to just five days this year. That counts for something.

Nonetheless, Amazon wants social media companies to take action and police platform activity without being prompted.

"While we appreciate that some social media companies have become much faster at responding, to address this problem at scale, it is imperative for social media companies to invest adequately in proactive controls to detect and enforce fake reviews ahead of our reporting the issue to them," it said.

Amazon did not identify which social media companies it has in mind, but security researchers have observed that Facebook and its Instagram subsidiary play a major role in the creation and distribution of fake reviews, as they do with fake news.

In an a preprint paper [PDF] distributed through ArXiv earlier this year, "The Fault in the Stars: Understanding the Underground Market of Amazon Reviews," Rajvardhan Oak, a graduate student at UC Berkeley who subsequently joined Microsoft, wrote, "Our research reveals that Facebook plays an important role in the continued operation of the underground reviews market."

Facebook does so using ads. "Targeted advertisements facilitate identifying users who might be complicit with writing incentivized reviews, and gradually sucks them deeper into the underground market," Oak's paper explains.

This approach, described in a 2019 research paper [PDF] as a "crowdturfing attack," can also be done through neural language models like GPT-2 NLM.

"Since this method has an economic cost, it is typically limited to large-scale attacks," the machine learning paper explains. "Automated crowdturfing, in which machine learning algorithms are used to generate fake reviews, is a less expensive and more efficient way to attack online review systems."

With Facebook reporting that it removed 1.3bn fake accounts in Q1 2021, down from 1.7bn in Q1 2020, perhaps there's some progress being made. But Facebook does not provide details about how long those accounts were active or how they were used during that time. To judge by Amazon's own seller forums, fake reviews continue to be a problem.

Thus Amazon has asked for "coordinated assistance from consumer protection regulators around the world."

Sonos sings

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

A Sonos executive [PDF] meanwhile decried anti-competitive big tech dominance, pointing to its inability to run Google Assistant on its speakers concurrently with other assistant software because Google would not allow it.

There were also academics who urged that something be done, without being overly certain about specific remedies.

"Today, the platform firms appear to many as an imperial power," observed Matt Crawford, Research Fellow at The Institute For Advanced Studies In Culture, in prepared testimony [PDF]. "The fundamental question 'who rules?' is pressed upon this body once again."

That's a provocative question in a room full of rule makers, some of whom differ over who legitimately rules the US. But it's one worth asking. ®

Broader topics


Other stories you might like

  • DuckDuckGo tries to explain why its browsers won't block Microsoft ad trackers
    Meanwhile, Tails 5.0 users told to stop what they're doing over Firefox flaw

    DuckDuckGo promises privacy to users of its Android, iOS browsers, and macOS browsers – yet it allows certain data to flow from third-party websites to Microsoft-owned services.

    Security researcher Zach Edwards recently conducted an audit of DuckDuckGo's mobile browsers and found that, contrary to expectations, they do not block Meta's Workplace domain, for example, from sending information to Microsoft's Bing and LinkedIn domains. Specifically, DuckDuckGo's software didn't stop Microsoft's trackers on the Workplace page from blabbing information about the user to Bing and LinkedIn for tailored advertising purposes. Other trackers, such as Google's, are blocked.

    "I tested the DuckDuckGo so-called private browser for both iOS and Android, yet neither version blocked data transfers to Microsoft's Linkedin + Bing ads while viewing Facebook's workplace[.]com homepage," Edwards explained in a Twitter thread.

    Continue reading
  • Despite 'key' partnership with AWS, Meta taps up Microsoft Azure for AI work
    Someone got Zuck'd

    Meta’s AI business unit set up shop in Microsoft Azure this week and announced a strategic partnership it says will advance PyTorch development on the public cloud.

    The deal [PDF] will see Mark Zuckerberg’s umbrella company deploy machine-learning workloads on thousands of Nvidia GPUs running in Azure. While a win for Microsoft, the partnership calls in to question just how strong Meta’s commitment to Amazon Web Services (AWS) really is.

    Back in those long-gone days of December, Meta named AWS as its “key long-term strategic cloud provider." As part of that, Meta promised that if it bought any companies that used AWS, it would continue to support their use of Amazon's cloud, rather than force them off into its own private datacenters. The pact also included a vow to expand Meta’s consumption of Amazon’s cloud-based compute, storage, database, and security services.

    Continue reading
  • Atos pushes out HPC cloud services based on Nimbix tech
    Moore's Law got you down? Throw everything at the problem! Quantum, AI, cloud...

    IT services biz Atos has introduced a suite of cloud-based high-performance computing (HPC) services, based around technology gained from its purchase of cloud provider Nimbix last year.

    The Nimbix Supercomputing Suite is described by Atos as a set of flexible and secure HPC solutions available as a service. It includes access to HPC, AI, and quantum computing resources, according to the services company.

    In addition to the existing Nimbix HPC products, the updated portfolio includes a new federated supercomputing-as-a-service platform and a dedicated bare-metal service based on Atos BullSequana supercomputer hardware.

    Continue reading

Biting the hand that feeds IT © 1998–2022