The Federal Bureau of Investigation (FBI) says cybercrime enabled by artificial intelligence resulted in losses exceeding $893 million in 2025, according to its latest Internet Crime Report.
The report, published by the Internet Crime Complaint Center (IC3), shows a sharp rise in the use of AI tools by criminals to execute increasingly sophisticated online fraud schemes.
It recorded 22,364 complaints linked to AI-enabled cyber activity during the year, with total adjusted losses estimated at over $893.3 million.
According to the FBI, criminals are using AI to generate convincing fake identities, including social media profiles, voice recordings and video content, making scams harder to detect.
Business email compromise schemes using AI-generated messages and voice cloning accounted for more than $30 million in losses, while impersonation and trust-building scams led to over $19 million in reported losses. Fraud involving cloned voices of family members in distress scams accounted for more than $5 million.
The report also highlighted large-scale losses from AI-assisted investment scams, which exceeded $632 million, representing the biggest share of AI-related fraud. In addition, AI-generated deepfakes and voice spoofing were used in recruitment and interview scams that caused nearly $13 million in losses.
The FBI warned that the growing accessibility of AI tools is expected to further accelerate cybercriminal activity, increasing risks for individuals, businesses and financial institutions globally.
