Introducing Daisy, the AI-generated grandma created by British telco Virgin Media O2 to be the ultimate con artist. Daisy’s only purpose is to talk to the imposters all day long so that real humans don’t have to.
Daisy debuted on November 14th and has had more than 1,000 conversations with scammers so far, the longest lasting about 40 minutes, but her lack of technology literacy has frustrated the scammers and caused her to lose her grandchildren. You’re wasting your time talking about irrelevant things.
“Daisy, a new member of our anti-fraud team, is turning the tables on fraudsters, outwitting them in a cruel game that only keeps them at risk.” said Murray McKenzie, director of fraud at Virgin Media O2. In a statement.
Daisy may sound like a human, but she’s essentially an AI large language model adapted from Grandma’s personality. She works by listening to the scammer and translating their voice into text. The AI then searches a large database to find an appropriate response based on the specific deception training it received and translates that text response into speech for Daisy to respond. All this happens in seconds without any additional input.
The company told CBS it worked with known scammers to train Daisy and put Daisy’s phone number on an online “mug list” (a list used by scammers targeting British consumers). He said he used a tactic called number seeding to get added. When they call, Daisy ends up “doing time” with the scammers.
Virgin Media O2 wants to draw attention to scams by not only wasting scammers’ time, but also by warning consumers to be wary of who they talk to.
“Daisy is a reminder that no matter how convincing the person on the other end of the phone, they are not always who you think they are,” McKenzie said.
According to the company’s research, 69% of British citizens report being targeted by fraud, and the company intercepted more than £250 million ($260 million) in suspected fraudulent transactions last year. It is said to have been prevented.
“Fraud has reached epidemic proportions, with organized groups of fraudsters operating professional call centres, relentlessly targeting Britons every second of every day,” said Virgin Media O2 Chief Operating Officer. (COO) Rob Orr said.
“We are constantly tightening our defenses and sharing compelling evidence about what these gangs are up to, but without real deterrence, these criminals are immune from the threat of prosecution. You can get away with it again and again.”
Fraud is also a big problem in the United States. According to caller ID and call blocking app TrueCaller, Americans receive an average of 2.9 billion unsolicited or spam calls per month. This equates to approximately 8 spam calls per user per month.
Although there is no U.S. version of Daisy, many U.S. phone companies have call screening software that can identify potential scams and robocalls.
On November 13, Google announced a new AI feature for Pixel smartphones that goes a step further and intercepts calls. Google AI sounds an alarm or vibrates when it detects conversation patterns commonly associated with fraud, such as a caller urgently requesting money or asking for account login or bank details. Masu.
As telcos and scammers continue to innovate with AI for their own purposes, the best thing consumers can do is remain vigilant.
An O2 spokesperson told CBS: “In terms of tips and tricks, it’s important to stay alert at all times, whether it’s shopping online, checking your email or receiving a surprise phone call.” he said. “We never call our customers and ask for their full password, one-time passcode, or bank account details over the phone, so calls like this are guaranteed to be a scam. Always check the links you receive and if something doesn’t seem true, it probably is.
contributed to this report.
