Sydney start-up Apate to change how telcos deal with scammers

An AI chatbot developed to waste a scammer’s time giving nonsensical answers is on the brink of changing how Australian telcos deal with scammers.

Apate founder Professor Dali Kaafar, centre, with his team.

An Australian start-up founded out of Macquarie University in Sydney is on the brink of changing how telcos deal with scammers, by using artificial intelligence to outfox them.

Telcos have for over a decade used technology to identify scammers based on their location and the type of device they are calling on – and automatically block them.

But Apate, founded by Professor Dali Kaafar, was built on the premise that one of the best ways to beat a scammer was to actually pick up their call and keep them on the line as long as possible, removing time they would have otherwise used on victims.

Now that approach is being considered by two major Australian telcos, including the nation’s third largest, TPG.

“Their current approach has been a very reactive approach. Service providers are obviously blocking a lot and they’re filtering these calls but as soon as they do that what happens is scammers use technology to spoof and place another 5000 calls,” Prof Kaafar said.

“The idea with Apate is to obviously distract them and have them diverted to bots.”

The shift away from what was initially planned to be a pure consumer app allowing people to pass a call to Apate, has now opened up another avenue for the start-up to do business.

Executive director of Macquarie University's cyber security hub, Dali Kaafar.

“Through the telco product, this is where the magic happens, if you like, because not only are you wasting a scammer’s time and you’re keeping them on the line, we’ve been able to develop a second product called Insights,” Prof Kaafar said.

“We’re extracting all sorts of intelligence from scammers from those particular conversations. That’s very powerful because currently there are no tools to extract accurate and timely intelligence about scam campaigns and what their tactics are.”

The new product arrives as the start-up has found a new form of phone scam is on the rise. Those are impersonation scams, when bad actors trawl through a company’s public information, pick the most accessible staff member and attempt to impersonate them to fleece vulnerable people.

While scammers have used the faces of celebrities and well-known business leaders for years on platforms like Meta’s Facebook, these new phone scams involve the use of lower-level employees who have an online presence.

This type of scam, where bad actors pretend to be real employees, using their name, title and subsequent information about them in a bid to fleece money from innocent people, now accounts for a little more than a quarter of all phone scams recorded by Apate through its pilot programs in April. That’s up from 18 per cent in March.

Impersonation scams more broadly, where scammers pretend to be from a company without impersonating a current staff member, accounted for between 44 per cent to 47 per cent of phone scams in April.

“Scammers are not only are impersonating organisations but they’re also impersonating profiles within these organisations, calling up saying they are from HSBC and pretending to a person who works there,” Prof Kaafar said.

“And if you go and look up those guys on Google or LinkedIn or whatever, you’ll find the people they are pretending to be.”

Apate, a new AI-powered bot designed to waste a scammer's time

Apate was formed more than 12 months ago during an afternoon picnic as he gave nonsensical answers to a scammer while his kids listened and laughed in the background. The name comes from Apate, the Greek god of deception.

When he brought the invention public, Apate had about 102 different personalities, including Mary, who had been designed to replicate a 65-year-old Australian woman.

She was among the top 10 at Apate, and on average kept scammers talking for more than 15 minutes before hanging up.

But now it has millions, and no one personality is ever the same, thanks to a $600,000 cash injection from the Office of National Intelligence, which Prof Kaafar and his team have used to further build out their technology.