Thieves Becoming More Sophisticated: Beware of 3 New Fraud Modus Operandi Draining Bank Accounts
Artificial Intelligence (AI) technology is making fraudulent tactics increasingly sophisticated and difficult to detect. Low-cost AI tools now provide widespread access to manipulate images and voices, as well as create websites that appear entirely official and legitimate.
Many AI-driven scams disguise themselves as ‘official’ advertisement videos on social media platforms like TikTok. For instance, a New York Times journalist reported encountering an advertisement for Hoka shoes offering an 80% discount on TikTok. Upon clicking, the user was directed to a site that appeared authentic; suspicion only arose when attempting to add items to the shopping basket. Investigations on Reddit revealed numerous other victims, prompting Hoka to issue warnings regarding these counterfeit sites.
This impersonation of legitimate websites is one of several advanced online fraud methods in the AI era. The FBI reported last month that cybercriminals defrauded US citizens of nearly US$21 billion last year, with approximately US$893 million of those losses linked to AI-related activities.
Essentially, AI facilitates the development of convincing fraudulent sites. Consequently, the public must adopt a more sceptical approach to avoid financial loss. “Instead of looking for indicators of what is bad, you now need to verify whether it is good,” said Mark Beare, General Manager of Malwarebytes, an internet security company.
The prevalence of fraudulent advertisements has led to increasing legal complaints against social media giant Meta. Last month, the American Consumer Federation filed a complaint accusing Meta of misleading users regarding its efforts to combat fraud, citing examples such as fake ads for baby supplies and free telephones. Similarly, Santa Clara County in California filed a lawsuit against Meta this month.
In response, Meta stated that it removed 159 million fraudulent ads and blocked nearly 11 million accounts affiliated with online scammers throughout the past year, while also increasing investment in anti-fraud technology. A TikTok spokesperson added that the platform prohibits fraudulent practices and misleading content, noting that 97% of rule-breaking spam content was removed in the fourth quarter of 2025 before being reported by users.
Beyond fake online stores, scammers are using AI to impersonate individuals, such as family members. The following are the three most common fraud methods in the AI era:
Video Call Scams: While traditional chat and phone-based scams are well-known, criminals are now using AI to manipulate video calls. They can use AI to swap faces in real-time to impersonate a victim’s acquaintance or relative, eventually leading to a request for money transfers. Andrew Yoon, a researcher at CivAI, noted that it is now very inexpensive to conduct real-time Zoom calls with highly realistic body and voice replacement.
Fake Celebrities: With the emergence of instant video generation tools like OpenAI’s Sora, social media is flooded with AI-generated fake content. Deepfake videos featuring Hollywood celebrities and business executives are widely circulated because vast amounts of their imagery are available to train AI models. For example, deepfake videos of chef Gordon Ramsay have been used to promote fake giveaways of cookware, tricking victims into providing credit card details under the guise of paying small shipping fees.
Fake Official Stores: AI-generated fraudulent shopping sites are rampant on social media, often targeting users based on their personal interests. Scammers purchase advertising space on TikTok and Instagram using the same targeting tools as legitimate marketers. Because they have no actual products to ship, they can spend more heavily on these targeted ads.
A simple way to identify fake online stores is to search the website’s URL on Google and check online forums for discussions from other victims. The most effective way to avoid online fraud remains simple: never trust advertisements that seem ‘too good to be true’ or video calls that suddenly request money. Stay vigilant.