Amazon’s using technology to fight against bad actors before its customers fall victim to Scamazon-style fraud and impersonation attempts.
As the holidays approach and last-minute shopping abounds, scammers prey on those trying to get in the spirit. Over the past several years, Amazon has amped up its game to prevent its customers from falling victim to those scams. And while the company’s efforts help consumers have a merrier holiday season, its fight against scams isn’t isolated to that time frame.
Related Stories
According to the Federal Trade Commission (FTC), impersonation scams were the most widely reported type of scam impacting U.S. consumers. Last year alone, consumers reported 330,000 instances of business impersonation scams and 160,000 instances of government impersonation scams. That figure doesn’t necessarily account for scams reported directly to an e-commerce company like Amazon or a third-party logistics provider like FedEx.
Scott Knapp, vice president of worldwide buyer risk prevention at Amazon, said though the company’s methods for putting the kibosh on reported scams have improved over time, the issue only continues.
“We know bad actors are getting better and better at mimicking anything that we can put out, in terms of warnings or authenticity,” Knapp told Sourcing Journal.
The company has begun to advise consumers to go directly to its app or site for any questions, concerns or issues with an order. Taking that step can help consumers identify when or if they have been scammed or spoofed.
The FTC noted in a blog post earlier this year that between 2020 and 2023, it saw a sharp decline in the number of phone calls perpetuating impersonation scams. Instead, consumers began reporting text message scams at much higher rates than before.
The agency also reported that multi-organization scams—that is to say, for instance, scammers pretending to be Amazon, then transferring victims to fake banks or fake government agencies for further investigation—has been on the rise.
Knapp said the ways scammers proliferate will only continue to change as artificial intelligence systems become more powerful and as the technology becomes more widely used. As AI grows, the company plans to use it to its advantage.
“Gone are the days of being able to have your suspicions raised just because the grammar is awful,” he said. “[This is] a result of bad actors having better and better tools to be able to imitate us or others. But on the other side of the equation, for all the AI they have, we’ve got it as well, and are using it on our side to detect…these things so we can warn customers about it.”
As part of the advanced AI and machine learning (ML) models Amazon employs, it has the capability to spot trends or detect emerging patterns in bad actors’ activity, which also gives the e-commerce and technology company a chance to alert its customers early on.
“Amazon uses AI and ML to find commonalities among all of the impersonation scam reports submitted by customers. From there, we can identify where to take action…and send [it] to law enforcement. AI effectively helps hold bad actors accountable at a greater speed,” Knapp explained.
The AI Amazon uses in its fights against scams and fraud doesn’t only extend to impersonation scams; it also helps the company uncover counterfeits and fake goods, widespread fraudulent reviews and sellers making consistent or material changes to existing goods on Amazon.
Knapp said tracking scam impersonation scam attempts can prove difficult because, unlike scams and fraud occurring on Amazon’s own platform, unless customers report them, Amazon has no record of impersonation scams happening.
But once the company can identify bad actors trying to harm its customers, action is near immediate. Amazon has the capability to remove scam phone numbers within the same business day and can see to the removal of phishing websites within hours, Knapp noted. In 2023, the company’s efforts resulted in about 40,000 phishing sites and 10,000 phone numbers being taken down because of impersonation scam reports it received and investigated.
But working in real time against scammers isn’t a task Amazon is interested in doing without support from other industry players.
“Even at our size, we don’t have the whole picture. So [by] working in consortiums and partnering with others, we can provide part of the data that then results in hundreds of call centers being taken down, but that probably wouldn’t happen if we were going it alone,” Knapp said. “That’s my big plea is, don’t be afraid to join…Any time we can stop fraud, we all get better; [there’s] no competitive advantage to going it alone against fraud.
