AI Chatbot Fools Scammers & Scores Money-Laundering Intel

Netcraft used ChatGPT-based chatbots over two years to pose as victims and engage scammers, successfully extracting banking details, account information across hundreds of institutions, and other operational intelligence about romance and investment fraud. The experiment demonstrates that AI personas (including local-language variants) can scale defender engagement, harvest otherwise hard-to-obtain infrastructure and payment data, and may precipitate AI-on-AI dynamics as attackers also adopt generative tools.