In July 2024, Amazon’s Prime Day event reaped an impressive £1.3bn from UK shoppers in just two days. It also led to a staggering 80% increase in impersonation scams, with around 120,000 fake Amazon websites reported during the same period. Amazon even had to warn Prime subscribers that they had “…noticed an increase in customers reporting fake emails,” as scammers attempted to trick people into handing over log in details.
It’s not just an Amazon problem. On Black Friday 2023, UK shoppers lost over £11.5m to online scams, marking a £1.1m increase from the year before, with the average loss per victim reaching roughly £695. Fraud tactics like identity theft also surged around the same period.
Richard Horne, the chief executive of GCHQ’s National Cyber Security Centre (NCSC) said fraudsters were “targeting bargain hunters with increasingly sophisticated scams, sometimes crafted using AI, making them harder to detect.” Another cybersecurity chief dubbed the event as “black fraud-day”, calling it “prime-time for cybercriminals.”
As these events stretch out across more and more days – Prime Day doubled in length from 2 days in 2024 to 4 days in 2025 – the window for fraudsters also widens. And when scams pan out successfully, consumers lose their money and the products they want, and retailers take a hit in exploited returns policies or false claims that packages weren’t delivered.
The Ugly Face of New AI Tools
With billions of pounds regularly changing hands amid these spending frenzies, fraudsters are embracing increasingly sophisticated strategies to exploit them.
While generative AI tools can simplify day-to-day tasks and unlock new efficiencies for businesses and customers alike, they also empower fraudsters with tools that are faster, smarter, and more scalable than ever before. Fraudsters are now using these technologies to create convincing phishing lures, product listings, payment links, fraudulent customer reviews, and more. With these tools at their disposal, fraudsters can exploit vast amounts of data with alarming ease.
Scammers can use AI to create fake websites that look eerily similar to the real thing, deceiving consumers into not only losing cash on items they’ll never receive, but handing over their login info as well. With deepfake technology, fraudsters can similarly impersonate celebrities, influencers, and even company executives. By hiding behind the credibility of an “established figure,” criminals can promote fake giveaways, investment schemes, or fraudulent product endorsements.
AI-powered bots are now automating fraud tactics like account takeovers or vulnerability exploitation, and can mimic the online behaviour of real human shoppers- making it increasingly difficult to distinguish between legitimate and malicious activity. What’s even more concerning is how widely these tools are being shared on the dark web. We’re seeing a surge in AI-generated fraud kits and plug-and-play tools that allow even novice fraudsters to experiment. In the past, getting into fraud required time, and mentorship. Now, with generative AI, anyone can get started with just a few clicks.
Between April and September last year, it was reported that retailers experienced over half-a-million AI-driven fraud attempts per day, including DDoS attacks.
Merchants Must Protect Themselves – And Customers
Merchants are stuck between a rock and a hard place.
Promotional discount events are indeed a great opportunity for retail merchants – last year’s £3.6bn in Black Friday and Cyber Monday sales nudged total UK Christmas spending over £22bn, while Boxing Day, the summer sales, and even back to school promotions contributed billions of pounds more to annual consumer spend.
On the one hand, shopping peaks like Black Friday and Amazon Prime Day are crucial for bringing in revenue and shoring up customer loyalty by offering deals.
On the other hand, the threat of AI-driven fraud is looming larger than ever and merchants must take action to protect their bottom line and consumers’ trust. What worked in the past may no longer be sufficient. Not only are fraud tactics becoming more sophisticated, but consumer behavior is also evolving. Shoppers are increasingly using generative AI to research products, get personalized recommendations, and even experiment with AI agents that find the best deals, introducing new patterns and vulnerabilities merchants must be ready for.
The first step businesses must take is integrating AI into their fraud stack. If they aren’t already leveraging dynamic, adaptive machine learning models, they’re operating at a serious disadvantage. These models allow merchants to assess risk in real time, dynamically adjusting policies to prevent abuse while maintaining a seamless customer experience. Once AI is in place, merchants can layer in additional protections, such as multi-factor authentication at login and checkout.
Advanced fraud solutions now connect device behavior, account activity, and historical transaction data across merchants to detect complex attacks and organized fraud rings. They also enable the analysis of transactional data to identify unusual patterns that may signal fraudulent activity. State-of-the-art machine learning can create identity and risk models that strengthen fraud prevention strategies without adding friction for good customers.
Companies like Rue Gilt Groupe are already adopting AI-driven fraud technology – not only reducing chargebacks, but also significantly improving approval rates. And with automated decisions, legitimate orders are processed within seconds, boosting both customer satisfaction and lifetime value.
Taking responsibility
With retail shopping peaks now becoming regular and familiar events in the calendar, retailers can better prepare for the risks.
Like Amazon did around the time of Amazon Prime Day, companies would do well to warn their customers about the risks of fraud and raise awareness on how to shop safely, use strong passwords, and identify fraudulent websites posing as authentic ones.
But retailers must go beyond simply warning their customers.Education is important, but it must be paired with implementation. By adopting fraud prevention technology that leans on data and AI-driven insights, retailers can distinguish bad actors without punishing loyal customers, making peak sales events more lucrative and safer for themselves and their valued customers.
