Bloomberg found out how AI helped scammers earn $442 billion

Residents of 42 countries lost $442 billion last year due to online scammers using artificial intelligence (AI), a sum that exceeds some estimates of the volume of the global cocaine trade, Bloomberg reports citing data from the Global Anti-Scam Alliance (GASA).

With AI, criminals create fake websites, imitate voices, and develop scenarios to defraud potential victims. These tools allow you to carry out phishing attacks and investment scams without special technical knowledge.

According to the company ZeroFox, which prepared a study for Bloomberg, in two years, the number of phishing pages found on free versions of nine platforms for developing and creating applications using AI has more than quadrupled to 10.6 thousand. The attackers created fake websites of banks, shops and telecommunications companies.

“The novelty of the situation is that AI-based code generators allow a person without programming skills to create a convincing fake login page for a banking service in just a few minutes, place it on a free subdomain and start collecting credentials even before it is detected,” ZeroFox said.

Along with AI, fraudsters use underground online platforms where you can purchase stolen personal data, fake accounts and software to create deepfakes. Money laundering services are also available, allowing criminals to set up schemes without their own sophisticated infrastructure.

The fight against fraud is complicated by the speed at which criminals create new sites and accounts. At the same time, financial transactions of attackers often pass through several countries, which makes it difficult to investigate and return stolen funds.

In Russia, fraudsters have also become more active in using artificial intelligence. According to the Center for Analytics of Social Engineering and Artificial Intelligence in Cyber Fraud, since the beginning of 2026, the number of deepfakes has doubled compared to the same period last year.

Attackers create fake videos and voices, including mimicking familiar people and famous faces, to convince users to transfer money or click on fraudulent links. Assistance in creating appropriate scripts is provided by specialized chatbots from the “dark” segment of AI.