BlockBeats reports that Australian AI anti-fraud company Apate.ai says it uses agents to collect scam intelligence

Article

AI A new angle on anti-fraud: proactively engaging with scams to obtain scam intelligence for regulatory agencies to block

According to BlockBeats,October 17, Australian AI anti-fraud startup Apate.ai (originating from research at Macquarie University in Sydney, 2024 established in ) has recently drawn attention. The company has deployed thousands to approximately 350000 AI agents, posing as ordinary victims of different ages and accents to answer scam calls, reply to phishing text messages, click malicious links, and infiltrate scammers' group chats around the clock.

Its core mission is to play dumb and delay scammers, extract intelligence such as scammers' bank account numbers, crypto wallets, and fake websites, and package it for banks and telecommunications companies to block. The company says it has completed more than 2500000 conversations, occupying scammers for a cumulative approximately 17 years of working time and extracting more than 250000 actionable intelligence items.

The company's clients include Commonwealth Bank of Australia and TPG Telecom, among others, with some institutions publicly reporting a significant decline in scam losses. 2026 August, Apate.ai completed a 8150000 USD seed funding round (Lobby Capital led the round), bringing total funding to nearly 10000000 USD, and is advancing global expansion, including establishing a London office and strengthening its US presence.

Original link https://m.theblockbeats.info/flash/372342

Key points

According to BlockBeats, Australian AI anti-fraud startup Apate.ai originated from research at Macquarie University in Sydney and was established in 2024.
Apate.ai has deployed thousands to approximately 350000 AI agents, posing as ordinary victims of different ages and accents, to answer scam calls, reply to phishing text messages, click malicious links, and infiltrate scammers' group chats.
The company says it has completed more than 2500000 conversations, occupying scammers for approximately 17 years of work time, and extracting more than 250000 actionable intelligence items.
The company's clients include Commonwealth Bank of Australia and TPG Telecom; some institutions have publicly reported a significant decline in scam losses.
2026 August, Apate.ai completed a 8150000 USD seed funding round,Lobby Capital led the round; total funding has reached nearly 10000000 USD, and the company is advancing its London office and US presence.

AI-assisted interpretation

The following is analysis, separate from reported facts. Verify important claims independently.

This is a “proactive response” anti-fraud approach:AI agents pretend to be real victims, interact with scammers and delay them, seeking to collect leads such as bank account numbers, crypto wallets, and fake websites, then provide them to banks and telecommunications companies for blocking.

Why it matters to readers

If the relevant data and results can be verified, this approach may help financial institutions and telecommunications companies identify and block scam infrastructure earlier.

Rather than asking users to identify scams, it has AI act on victims' behalf to engage scammers and collect leads for institutional handling.

Risks and unknowns

Information on the number of deployed agents, conversation volume, intelligence volume, and funding primarily comes from company statements or media reports, with no independently verified supporting materials provided.
The specific institutions, time periods, and decline figures behind the claim of a “significant decline in scam losses” have not been disclosed.
The evidence does not explain which blocking outcomes were actually caused by the collected intelligence.

BlockBeats reports that Australian AI anti-fraud company Apate.ai says it uses agents to collect scam intelligence

Article

AI-assisted interpretation

Why it matters to readers

Risks and unknowns

Related Developments

Related concepts

Keep following

Security Center