Automating Deception: AI s Evolving Role in Romance Fraud
The increasing sophistication of AI-based tools is transforming the landscape of online deception. Fraudsters have long relied on pre-scripted messages and structured manipulation techniques to exploit victims, but AI offers new opportunities to scale and refine these operations. Generative AI has already been exploited for various cyber-enabled crimes, and recent studies indicate that there is a growing interest in leveraging AI for social engineering and financial fraud. The development of industrialised romance fraud, as seen in large-scale scam networks in West Africa and Southeast Asia, indicates that AI could not only enable new fraudsters but further streamline existing operations. Yet while LLMs excel at identifying the procedural elements of romance fraud, they still struggle to sustain these deceptive patterns over extended interactions. This asymmetry creates a technical opportunity for defensive applications, particularly in detecting AI-generated content across the romance fraud lifecycle.