TPG, an Australian telecom provider, has routed nearly 590,000 calls to Apate’s scam-engagement service; the Australian startup closed an $8.15 million seed financing in August, with total funding nearing $10 million.
Agents use invented personas spanning different ages, genders, accents and languages. Their routines include taking scam calls, answering suspicious messages, clicking phishing links and joining scam groups. They log bank details alongside crypto wallets, phone numbers and malicious sites. Customers can use these leads to stop funds reaching scammers before victims lose money.
Agents have logged 2.5 million conversations with suspected scammers. Their recorded interactions add up to 17 years.
One bank customer applied Apate information to break up 22 scam operations. Commonwealth Bank is also a customer; its scam losses declined 76%.
Deepfakes of executives, public figures and other trusted people are one way scammers use AI; automated conversations are another. Both methods reduce the expense of approaching prospective victims. Apate’s system is intended to keep contact going and build a record of accounts and other infrastructure involved in scams.
Research at Macquarie University in Sydney provided the basis for Apate, founded in 2024. CEO Dali Kaafar’s idea came after he kept a scam caller on the line for 44 minutes at a family picnic, then considered whether software could manage similar interactions.
