While most businesses currently deploy artificial intelligence to increase productivity or automate workflows, the startup Apate.ai has adopted a different approach. The company is programming AI agents to act as potential victims, intentionally entering into conversations with scammers to study their methods. By going 'undercover' in these interactions, the AI collects data on how fraud is perpetrated in real-time.

The primary objective of this project is to build a repository of fraud intelligence. By observing the techniques, scripts, and psychological tactics used by bad actors, Apate.ai aims to develop more effective defensive tools for companies and consumers. This proactive strategy shifts the role of AI from a passive defensive shield to an active participant in gathering threat data. The company suggests that by wasting the time of scammers and documenting their patterns, they can better understand the evolving landscape of digital fraud. No major disagreements between industry analysts have emerged regarding the efficacy of this method, though the practice of using AI to engage with criminals remains a novel and experimental approach in the cybersecurity sector.