Australian startup Apate.ai uses thousands of bots disguised as possible scam victims to draw out fraudsters and gather details about their operations. By the company’s count, its bots have held 2.5 million conversations, using up 17 years of scammers’ time in total.
Apate was established in 2024 after research at Macquarie University in Sydney. Its agents take on identities with varying ages, genders, accents and languages. They answer scam calls, respond to suspicious emails and messages, open phishing links, and enter social media groups to lure fraudsters into conversations. Keeping scammers engaged gives the agents time to seek useful details.
The information they seek includes bank accounts, crypto wallets, phone numbers, malicious websites, brands being impersonated and other specifics about a scam. Such details typically come to light only after someone has suffered a loss. Banks, telecom companies, government agencies and utilities receive Apate’s findings and use them to guide their anti-fraud measures and initiatives.
“Most of the industry works after the fact. Telcos block numbers, but scammers just get new ones. Banks watch transactions, but by then the scammer has already reached the customer. We go to the scammer directly and get the information before anyone gets hurt,” Apate co-founder and chief commercial officer Brad Joffe told TechCrunch.
“Our AI agents talk to tens of thousands of scammers every day across calls, messages, email and the web. What we offer our customers is what comes out of those conversations, and what they can do with it.”
Customers can give an agent a specific task, including finding a bank account number or a crypto-wallet address. The text agents can scan forums, investment groups and social platforms for possible scam activity, then join a discussion under a made-up identity to draw out information about the scammers.
Customers also have access to an insights portal alongside the agents. It provides intelligence and analytics about scams aimed at their end users or employees, brands, products and more.
To operate the agents and analyze what they collect, Apate uses both frontier models and its own fine-tuned models, Joffe said. The company trains them using conversations between its bots and scammers, together with founder and CEO Dali Kaafar’s research at Macquarie.
Commonwealth Bank of Australia and TPG Telecom are among Apate’s named customers; it also serves other utilities, telcos and government agencies. Joffe said nearly 590,000 calls from TPG have been redirected to the bots, while publicly reported scam losses at Commonwealth have dropped by 76%. An unnamed Australian bank that uses the service disrupted 22 scam campaigns with Apate’s outbound agents, he said.
AI is also a tool for fraudsters. The FBI received more than 22,000 complaints about AI-related crimes in 2025, representing about $893 million in losses. Of that amount, more than $632 million involved investment scams. According to the FBI, fraudsters running those scams use AI for conversations and to create deepfakes of the voices and appearances of celebrities, executives and trusted figures.
People were baiting scammers well before these bots arrived. For years, hobbyists have kept fraudsters busy by, for example, convincing them to read whole books out loud or get absurd tattoos. As AI becomes more capable, Apate can put that approach on an automated footing and gather intelligence during the conversations.
A scam call to Kaafar set the idea in motion. While attending a family picnic several years ago, he received a call from a number he didn’t recognize. Once he realized it was a scammer, he feigned ignorance to amuse his children and kept the caller talking for 44 minutes.
The family laughed, but Kaafar also saw that the scammer hadn’t been able to call a real victim while talking to him. Those minutes had cost Kaafar time, too. That raised a question: could he automate the conversation and get useful information from it at the same time?
Kaafar pursued the question with colleagues at his workplace, Macquarie University. They applied machine learning and natural language processing to investigate what scam calls contained and look for better methods of detecting and preventing scams. That work ultimately became the foundation for Apate.
“Apate is named after the Greek goddess of deception, which felt appropriate given what our agents do,” Joffe said. Even the company’s pitch material features scammers swearing at the bots.
“We track that as a sign of frustration and every swear word gets celebrated by us at Apate. We are probably one of the only businesses in the world that celebrates when people swear at our product.”
In August, Apate raised $8.15 million in seed funding from Lobby Capital, OIF Ventures, Investible, Concept Ventures and Baobab Ventures. The company has 20 people and has raised just under $10 million overall, according to Joffe. He said the new funding will support work on the product and international expansion, with plans for a London office and a bigger presence in the U.S.
Apate is a finalist in this year’s Startup Battlefield 200 at TechCrunch Disrupt and will present its pitch to a panel of venture capital judges there. Disrupt is being held this week at San Francisco’s Moscone West, continuing through Thursday, October 15.
