AI Grandma Battles Scammers to Protect Consumers

• 25 Nov 2024

AI Grandma Battles Scammers to Protect Consumers

 

Virgin Media O2 has introduced Daisy, an AI-powered "grandma" designed to combat phone scams by wasting scammers' time, protecting real customers from the hassle. Debuting on November 14, Daisy’s primary function is to engage with scammers and keep them occupied, allowing people to avoid being targeted. In just a few weeks, Daisy has already conversed with over 1,000 scammers, with the longest interaction lasting nearly 40 minutes. Daisy frustrates fraudsters by pretending to be tech-illiterate and sharing irrelevant, long-winded stories about her grandchildren, ultimately leading the scammers nowhere.

 

Murray Mackenzie, director of fraud at Virgin Media O2, described Daisy as “the newest member of our fraud-prevention team," emphasizing how the AI outwits fraudsters by simply keeping them on the phone for as long as possible. Daisy may sound like a human, but she is powered by an AI large language model, with the persona of a kindly grandmother. The system works by transcribing the scammers' speech to text, analyzing the conversation, and then selecting an appropriate response based on training in fraud prevention. This process happens in real-time, with Daisy replying almost immediately, without any manual input needed.

 

Virgin Media O2 trained Daisy by working with real-life scammers and utilizing a technique called number seeding. Through this tactic, Daisy’s number was added to scammer databases, known as “mug lists,” which fraudsters use to target victims. Once scammers dial in, Daisy ensures that they stay on the line, using the opportunity to waste their time. Virgin Media O2 hopes that this initiative will not only disrupt scam calls but also raise awareness about fraud and encourage people to be more cautious when answering unknown calls.

 

Daisy also serves as a reminder to the public that phone scammers can be persuasive and convincing, but they are not always who they claim to be. According to Virgin Media O2's research, 69% of Brits have been targeted by scams, and the company has intercepted and blocked over £250 million ($260 million) in fraudulent transactions in the past year. "Fraud is at epidemic levels," said Rob Orr, Virgin Media O2’s COO, who explained that organized crime groups run professional scam operations that target the public relentlessly. Despite these efforts, the lack of deterrents allows these criminals to continue their operations with little fear of prosecution.

 

In the United States, fraud is also a significant issue. TrueCaller, a caller ID and blocking app, reports that Americans receive about 2.9 billion unwanted or spam calls every month—an average of eight per user. While there is no equivalent to Daisy in the U.S., many American phone companies have begun using AI-driven call-screening tools to identify suspected scams. For example, Google announced a new feature for its Pixel phone in mid-November that listens in on calls and alerts users if it detects common scam patterns, such as urgent requests for money or sensitive account information.

 

As phone companies and scammers continue to innovate using AI, consumers must remain vigilant. Virgin Media O2 encourages the public to follow a few simple rules: always verify calls that ask for sensitive information, avoid clicking on suspicious links, and trust your instincts—if something seems too good to be true, it likely is. A spokesperson from O2 added that the company will never call customers asking for full passwords, one-time codes, or bank details, so any call requesting such information is a sure sign of fraud.