Interpol: AI-based fraud generates 4.5 times |

✨ Megiddo

✨ President ✨
Staff member
912
246
Interpol estimates that AI increases fraudsters' profits by 4.5 times. Financial fraud losses in 2025 were estimated at $442 billion, and the organization expects this figure to grow, primarily due to artificial intelligence.

According to Interpol's annual Financial Fraud Report, AI significantly improves the efficiency of criminal schemes. The organization estimates that AI-enabled fraudulent operations generate 4.5 times more profit than conventional ones.

Criminals most often use generative AI to "polish" details that previously gave them away. For example, LLMs help rewrite phishing emails and messages to eliminate errors and stylistic nuances typical of non-native speakers. This is critical for attacks impersonating major brands, where any inaccuracy can arouse suspicion.

At a more advanced level, attackers use deepfakes. Interpol notes that deepfake technology has advanced significantly even in the past two years: criminals can now create convincing voice clones with as little as ten seconds of recording (for example, audio from a social media post).

Darknet marketplaces even sell ready-made kits for creating synthetic identities (these products are called deepfake-as-a-service), which allow victims to convince themselves they are communicating with a familiar person. Such kits are inexpensive, and law enforcement believes they are accelerating the "industrialization" of cybercrime.

The report also devotes special attention to agent-based AI. While this technology is not yet widely used, Interpol fears it is only a matter of time. For example, with the help of AI agents, fraudsters will be able to automate the collection of information on victims—from login credentials to vulnerabilities in company infrastructure.

Moreover, agents can analyze stolen data and, for example, suggest ransom demands to attackers based on the value of the stolen information and the victim's financial situation.

The organization has also recorded an increase in sextortion schemes using AI-generated images. In several cases studied, victims initially rejected traditional offers from scammers (cryptocurrency and forex scams, romance scams), but were subsequently blackmailed using deepfakes.

“Thanks to artificial intelligence, cheap digital tools and growing international cooperation between criminal networks, we are seeing the industrialization of fraud,” said Interpol Secretary General Valdecy Urquiza.