AI-powered deepfake identity thefts surge 3,000%

Several months ago, a worker in Hong Kong unexpectedly conducted a transaction worth $25 million for her company, following instructions given by her financial director and other high-ranking executives via teleconference. It later turned out to be an incredibly elaborate fraud scheme using deepfake technology, where audiovisual material of the company's leaders was used to "convince" the employee to transfer the huge sum.

This case, while shocking, is not isolated and is likely to repeat. In the past year, attempts to use fake identities through deepfake technology have surged, recording a 3,000% increase. This surge signifies the addition of a new and potent threat to financial frauds.

The use of stolen or fake identities is the prime method for gaining access to bank accounts, with victims worldwide experiencing cumulative losses in the billions. Now, fraudsters are leveraging the tools of productive artificial intelligence to mimic voices or the faces of others.

Reports are already surfacing of calls from alleged relatives and friends seeking financial aid and sounding just like their genuine selves. While they may not ask for large sums, experts advise caution, recommending to hang up and call the purported friend or relative directly if in doubt.

Fraudsters Favoring Targets Outside Europe
Australia has witnessed a surge in such activities, with entire news pieces and videos being created using deepfake technology, often featuring a celebrity, promoting supposed investment opportunities. The national fraud center was compelled to issue warnings, stating that such scams cost Australians at least $9 million last year.

Similar efforts were noted closer to home in Europe in October 2023 when videos surfaced mimicking BBC presenters allegedly promoting an investment program with Elon Musk. However, according to recent data, Europe is the least favored region for fraudsters.

In the United States, the FBI's specialized cybercrime center received 900,000 complaints last year, marking a 22% increase compared to the previous year. Losses may exceed $12.5 billion, a figure expected to rise by $2 billion annually in the future.

"Fraudsters are increasingly using genuine documents (acquired through data leaks) and then altering their faces for biometric checks," states a report by Onfido. While the forgery of biometric data is on the rise, it still constitutes a small portion of frauds. Conversely, there's an exponential increase in the use of productive AI and deepfake technology.

According to the same report, a small number of fraudsters are responsible for the majority of deepfake attacks, which tend to target one business at a time. Another trend is that attacks occur every day, not just on weekends as before, indicating that criminal activity is international and interconnected.