AI anti-scam tools steer fraudsters into long traps

AI is moving from protecting you after a scam starts to helping stop scams before they do real damage. A new generation of automated systems is designed to bait scammers, keep them talking, and quietly collect intelligence that can be used to shut down operations faster. That shift is happening now, with real-world pilots and growing interest from banks, telecoms, and security researchers.

One standout example is Apate, an Australian startup named after the goddess of deception. The company has built a platform that deploys hundreds of thousands of AI-powered “victims” across scam calls and chat groups. The idea is simple in theory: keep scammers talking long enough to waste their time, while gathering intelligence about the kinds of scams they run, the money mules they rely on, and the URLs they push. Apate says its bots can imitate different personas, switch languages, and appear skeptical to avoid tipping off scammers that they’re not talking to a real person. In practice, this means a scam call that might otherwise lead to a real loss can be dragged out, giving defenders more chances to intervene elsewhere in the system.

In tests run by media writers using a demo version of the platform, the AI-driven victims stayed engaged with scammers for long stretches, often more than two hours, while remaining resilient to simple “get rich quick” ploys. The goal isn’t to swindle the scammers but to burn time and harvest actionable data. And it’s not just calls. The bots also engage scammers in text chats and can participate in scam-related online groups to map the spread of fraud campaigns in real time.

Security researchers describe these efforts as a form of psyops or honeypot work. The idea is to lure criminals into controlled environments where their methods, tools, and targets can be studied safely. Experts say this kind of automation could help improve information sharing across industries, from banks to online platforms, and could even inform new takedown efforts against large scams operations.

That broader vision matters for everyday users in the United States, the United Kingdom, and Canada because it could translate into quicker responses when a scam trend starts, better warning signals on bank and credit-card apps, and sharper fraud rules built on real-world data rather than isolated incidents.