An Australian company says its AI bots keep phone and online scammers talking while collecting information about their operations. The approach turns conversational AI towards disrupting fraud, not just detecting it after the damage is done.
Watch Desk analysis
What happened
Apate says its system diverts phone scammers to AI-generated callers designed to keep conversations going without falling for the scam. The bots also join scam chat groups and respond to messages. Company founder and chief executive Dali Kaafar says the platform, used by banks and supported by telecoms companies, has around 350,000 bots and has collected more than 250,000 pieces of information about fraudsters, including scam URLs and bank details.
The figures and platform description come from Apate. In a hands-on demonstration, WIRED tested Apate’s scam-baiting system, playing scammers trying to persuade its AI personas to invest in cryptocurrency. The publication found the conversations responsive and natural, but the demonstration was not a test of real-world fraud reduction.
Why it matters
Scam operations can make enormous numbers of calls and messages. Keeping a fraudster occupied may reduce the time they have for potential victims, while the conversation could reveal details useful to banks, telecoms firms or investigators. It is an intriguing change of direction: the bot is not trying to spot a scammer from a safe distance, but to become the person the scammer thinks they have hooked.
There is a related idea in cybersecurity honeypots, decoy systems used to attract attackers and learn how they operate. ETH Zurich researcher Mark Vero says his team found that an LLM-powered honeypot kept AI agents attacking for significantly longer than a more predictable one, and that the agents were less likely to identify it as a honeypot. That is a research finding about simulated attacks, not proof that scam-baiting bots prevent fraud at scale.
Our read
This is AI with a job more specific than “make a convincing chatbot”: waste a criminal’s time and gather clues. Apate’s scale claims are striking, but the reader-facing test is whether the information leads to disrupted scams or fewer people being targeted. Making scammers talk is a promising tactic; it is not, by itself, a measure of protection.
What to watch
- Whether Apate or its partners publish evidence of scams disrupted or victims protected.
- How banks, telecoms firms and investigators use the information gathered by the bots.
- Whether AI-powered honeypots prove effective against real attackers, not only simulated ones.
Discussion spark: Should banks and telecoms use AI bots to engage scammers and gather intelligence, even if it remains unclear how much that reduces harm to potential victims?
Sources and evidence
- Source update (10 October 2026, 12:00 UTC)
Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.