Australian company Apate.ai uses thousands of AI agents to pose as potential victims of scams, then interact with scammers through phone calls, messages, email, phishing links and social media groups. The goal is not merely to waste scammers’ time, but to collect operational information that banks, telecommunications companies, government agencies and utilities can use to improve anti-fraud systems.
Apate says its fleet has conducted 2.5 million conversations to date, equivalent to wasting 17 years of scammers’ time. During these interactions, the agents attempt to uncover bank account numbers, cryptocurrency wallets, phone numbers, malicious websites and the brands impersonated by scammers, as well as details of the campaigns they run.
From wasting time to gathering evidence
The agents can assume personas that differ in age, gender, accent and language, helping them keep scammers in conversations for longer. Customers can also assign them specific tasks, such as obtaining a bank account number or cryptocurrency wallet address. The company provides customers with an analytics portal displaying information about scams targeting their users, employees, brands or products.
Apate relies on advanced models as well as others fine-tuned internally by the company, and trains its models on conversations that the agents conduct with scammers, alongside research carried out by founder and CEO Dali Kaafar at Macquarie University.
What changes in practice?
The approach focuses on shifting from post-incident response to collecting indicators before harm occurs. According to Brad Joffe, the company’s co-founder and chief commercial officer, telecommunications companies’ blocking of phone numbers does not prevent scammers from using new numbers, while monitoring banking transactions may happen after a scammer has reached a customer. Apate therefore attempts to obtain information directly from conversations with scammers.
Customers include Commonwealth Bank of Australia and TPG Telecom, along with utilities, telecommunications companies and other government agencies. The company said TPG routed around 590,000 calls to its agents, and that Commonwealth recorded a 76% decline in publicly reported fraud losses, while an unnamed Australian bank used the external agents to disrupt 22 scam campaigns.
Funding and international expansion
Apate was founded in 2024 based on research conducted at Macquarie University in Sydney. The idea for the company emerged after a scam call in which Kaafar kept the scammer on the line for 44 minutes, before the idea developed into research on machine learning and natural language processing to extract call content and improve fraud detection and prevention.
In August, the company raised $8.15 million in a seed funding round from Lobby Capital, OIF Ventures, Investible, Concept Ventures and Baobab Ventures. Joffe said total funding had approached $10 million, and that the company, which has 20 employees, intends to use the latest round to develop the product and expand internationally, including opening an office in London and strengthening its presence in the United States.
The service comes at a time when the source noted that the U.S. Federal Bureau of Investigation received more than 22,000 complaints related to crimes involving artificial intelligence during 2025, with losses approaching $893 million, including more than $632 million linked to investment scams. The reported results concerning reduced losses and disrupted campaigns remain claims made by the company, and the source did not provide independent details about the measurement methodology or mechanisms for handling the legal and operational risks of automated interaction with scammers.