Cybercriminals are shifting from large fraud compounds to smaller groups using AI, making scams harder to detect, warned Giles Thomson, president of the Financial Action Task Force. Fraudsters previously operated in massive buildings with hundreds of workers, but now use one or two individuals with powerful servers and AI tools. Criminals in south-east Asia develop trust-based relationships with targets before deploying cryptocurrency hoaxes, investment frauds, and romance scams.

AI enables more complex schemes like deepfake videos, fake websites, and chatbot impersonation. Interpol found AI-enabled fraud is 4.5 times more profitable than traditional methods, with a 54% rise in fraud reports since 2024. The FBI reported $893mn in losses from AI-driven fraud last year. An upcoming FATF report may recommend anti-scam centres for faster information sharing. Myanmar faces potential extra countermeasures over its extensive fraud operations.

The US Department of Justice recently shut down 13 Chinese-run compounds in Madagascar, freezing $52mn in crypto linked to Xinbi Guarantee. Global financial fraud losses reached $442bn last year, per the Global Anti-Scam Alliance.