UK AI Grandma Daisy Wastes Scammers’ Time, Raising Fraud Awareness

Daisy, an AI-generated “grandma” developed by British phone company Virgin Media O2, is taking the fight against scammers to a new level. Her sole mission? Waste scammers’ time so real people don’t have to.

Since her launch on November 14, Daisy has already engaged in over 1,000 conversations with scammers. Some of these calls have lasted up to 40 minutes, with Daisy frustrating fraudsters by feigning tech-illiteracy and sharing irrelevant stories about her grandchildren.

According to Murray Mackenzie, director of fraud at Virgin Media O2, Daisy is effectively “turning the tables on scammers—outsmarting and outmaneuvering them at their own cruel game simply by keeping them on the line.”

How Daisy Works

While Daisy may sound like a chatty human, she’s powered by an AI large language model designed specifically for scam prevention. Her process is simple but effective:

• Voice-to-Text Translation: Daisy listens to the scammer’s voice and converts it to text.

• Scam-Specific Training: Based on her training and database of known scams, Daisy identifies the scam type and generates a suitable response.

• Text-to-Speech Conversion: Daisy replies with her distinct grandma-like voice, keeping the scammer engaged.

• This entire process happens in seconds, with no human intervention required.

To ensure scammers call Daisy, Virgin Media O2 used a tactic called number seeding, which places Daisy’s phone number on “mugs lists”—databases of potential scam targets used by U.K. fraudsters. This ensures Daisy gets plenty of calls to keep scammers busy.

Raising Awareness About Fraud

Daisy isn’t just about wasting scammers’ time; Virgin Media O2 also hopes to raise awareness about the growing threat of fraud. Mackenzie reminds consumers:

"Daisy is also a reminder that no matter how persuasive someone on the other end of the phone may be, they aren't always who you think they are."

Virgin Media O2’s research reveals that 69% of Brits have been targeted by scams, and the company intercepted over £250 million ($260 million) in fraudulent transactions last year. However, fraud remains a significant issue, with organized crime groups running professional call centers to target consumers relentlessly.

The Fight Against Scams: A Broader Perspective

Fraud is not just a U.K. issue; it’s a global problem. In the U.S., for example, Americans receive an average of 2.9 billion spam calls per month, equating to eight spam calls per user, per month, according to TrueCaller.

While there’s no American equivalent to Daisy, U.S. phone companies offer call screening software that flags potential scams. Google recently announced a new AI feature for its Pixel phones, which analyzes calls for scam-like behavior, such as urgent money requests or attempts to obtain sensitive information.

Staying Vigilant

As scammers and phone companies continue to innovate with AI, the responsibility of staying vigilant falls on consumers. Virgin Media O2 offers the following tips to avoid falling victim to scams:

• Never share full passwords, one-time passcodes, or bank details over the phone.

• Be cautious of links sent via text or email; always verify their legitimacy before clicking.

• If something sounds too good to be true, it probably is.

What This Means

Daisy highlights how AI can be leveraged not just for convenience but for societal benefit. By keeping scammers occupied, she protects potential victims and raises awareness about the ever-growing threat of fraud.

As fraud prevention tools like Daisy and Google’s AI advance, the hope is that scammers will face increasing challenges in targeting consumers. Until then, vigilance remains the best defense

Editor’s Note: This article was created by Alicia Shapiro, CMO of AiNews.com, with writing, image, and idea-generation support from ChatGPT, an AI assistant. However, the final perspective and editorial choices are solely Alicia Shapiro’s. Special thanks to ChatGPT for assistance with research and editorial support in crafting this article.