Apate deploys undercover AI agents to bait fraudsters
Australian cybersecurity startup Apate is deploying thousands of persona-driven AI agents to impersonate scam targets, helping institutions gather actionable threat intelligence directly from fraudsters.
Apate, an Australian security startup spun out of Macquarie University research, deploys thousands of AI agents to actively engage fraudsters across phone calls, text messages, emails, and social media platforms. By masquerading as targets across diverse ages, accents, and languages, these agents keep scammers engaged to collect crucial intelligence—including crypto wallet addresses, bank accounts, and malicious domains—before consumers fall victim.
To date, Apate claims its fleet has engaged in 2.5 million conversations, wasting 17 years of scammers' time. The company works with enterprise clients like Commonwealth Bank of Australia, TPG Telecom, utilities, and government agencies. TPG Telecom has diverted nearly 590,000 calls to the platform, while Commonwealth Bank reported a 76 percent drop in publicly reported scam losses. Another Australian financial institution leveraged outbound agents to disrupt 22 scam campaigns.
The platform relies on a combination of frontier models and fine-tuned proprietary models trained on academic research and real-world interactions. Clients can assign specific missions to agents, such as extracting payment credentials, while monitoring threat analytics through a central insights portal. The expansion comes as automated crime scales rapidly; the FBI recorded over 22,000 complaints regarding AI-related crimes resulting in $893 million in total losses, with $632 million originating from investment scams.
To fund international expansion into London and the U.S., the 20-person startup raised an $8.15 million seed round led by Lobby Capital, OIF Ventures, Investible, Concept Ventures, and Baobab Ventures, pushing total funding to nearly $10 million. Co-founder Brad Joffe emphasized that rather than blocking numbers after the fact, "We go to the scammer directly and get the information before anyone gets hurt." For security practitioners, this approach offers proactive threat intelligence that turns scam interaction into automated counter-defensive research.
This is our own summary of reporting by TechCrunch AI
