An Australian startup is deploying artificial-intelligence agents to impersonate potential scam victims—not to transfer money, but to keep fraudsters talking long enough to expose their infrastructure. TechCrunch reported on October 14 that Apate operates thousands of decoy agents across calls, text messages, email, phishing links and online groups.
Founded in 2024 from research conducted at Macquarie University in Sydney, Apate gives its agents fabricated identities spanning different ages, genders, accents and languages. The bots are designed to appear vulnerable while sustaining conversations that might otherwise target real people.
The company says its agents have participated in 2.5 million conversations and collectively consumed 17 years of scammers’ time. Apate also tries to extract information that investigators often receive only after a victim has lost money, including bank-account details, cryptocurrency-wallet addresses, phone numbers, malicious websites, impersonated brands and descriptions of the underlying scheme. Those figures and capabilities are company claims reported by TechCrunch.
Apate passes the resulting intelligence to customers including banks, telecommunications providers, utilities and government agencies. Customers receive an analytics portal focused on scams aimed at their users, employees, brands or products. They can also assign agents specific missions, such as obtaining an account number or wallet address, while text-based agents search forums, investment groups and social platforms for possible fraud activity before entering conversations under fictional identities.
The need is substantial. TechCrunch cited FBI data showing more than 22,000 complaints involving AI-related crimes in 2025 and about $893 million in associated losses. Investment scams accounted for more than $632 million. The FBI says fraudsters use AI to sustain conversations and to create deepfakes of celebrities, executives and other trusted figures.
Apate grew from a scam call received by founder and chief executive Dali Kaafar during a family picnic. Kaafar kept the caller occupied for 44 minutes, then asked whether the distraction could be automated and turned into useful intelligence. His subsequent research at Macquarie used machine learning and natural-language processing to analyze scam-call content and improve detection and prevention methods.
The startup says it combines frontier models with its own fine-tuned systems, training them on conversations collected by its agents and on Kaafar’s university research. Co-founder Brad Joffe told TechCrunch that the aim is to approach fraudsters directly and gather actionable information before banks or telecom providers are left responding to a completed scam.
Apate named Commonwealth Bank of Australia and TPG Telecom among its customers. Joffe said TPG had diverted nearly 590,000 calls to the agents, Commonwealth had seen a 76% reduction in publicly reported scam losses, and an unnamed Australian bank used outbound agents to disrupt 22 campaigns. TechCrunch did not report independent verification of those outcomes or establish how much of Commonwealth’s decline was caused by Apate.
The 20-person company raised $8.15 million in seed funding in August from Lobby Capital, OIF Ventures, Investible, Concept Ventures and Baobab Ventures, bringing total funding to just under $10 million, according to Joffe. Apate plans to use the money for product development and international expansion, including a London office and a larger U.S. presence. The report does not detail the agents’ error rates, legal boundaries across jurisdictions, or safeguards governing their use of fabricated identities, all of which will matter as the system expands.
Comments
Loading comments…