Australian startup Apate says its AI agents have handled millions of scam interactions and supply intelligence to banks, telecoms and government clients.
Australian startup Apate uses conversational AI agents to pose as potential victims, engage suspected scammers and gather information about their operations, according to a TechCrunch report published October 14. The agents can respond to calls, messages and emails, follow phishing links and enter online groups, aiming to keep fraudsters occupied while collecting details such as phone numbers, bank accounts, crypto wallets and impersonated brands.
Apate says its agents have taken part in 2.5 million conversations and collectively wasted 17 years of scammers’ time. The company, founded in 2024 following research at Macquarie University, says it serves banks, telecom companies, government agencies and utilities. Its named customers include Commonwealth Bank of Australia and TPG Telecom. Apate’s co-founder told TechCrunch that TPG had diverted nearly 590,000 calls to the agents; he also said Commonwealth Bank had seen a 76% decline in publicly reported scam losses. Those outcome figures are company-reported, and the article does not establish that the service alone caused the reduction.
The approach differs from conventional defenses that focus on blocking known numbers or flagging transactions after fraud attempts. Apate instead seeks to interact with scammers directly and pass intelligence from those exchanges to customers for use in prevention and investigation. The company said one unnamed Australian bank used its outbound agents to disrupt 22 scam campaigns. It combines frontier models with its own fine-tuned models, according to the report.
Apate raised $8.15 million in seed funding in August and said it had raised just under $10 million in total. The company plans to use the money for product development and international expansion, including a London office and a larger US presence. The model’s practical test will be whether clients can turn the gathered intelligence into measurable reductions in fraud without creating new risks in how the agents interact with people or handle information.