Inspiration from a story by Rob Thubron on Techspot, November 2024.
The rise of human-like AI has sparked valid concerns about its potential to displace workers, but one company has turned this technology toward combating a global nuisance: phone scammers. Fighting fraud with AI, they’ve created a tool designed to mimic a talkative senior citizen, engaging fraudsters in endless conversations akin to the rambling tales of Grandpa Simpson.
The innovative AI, nicknamed "Daisy" (stylized as dAIsy), was developed by O2, the UK’s leading mobile network operator. Daisy convincingly poses as a grandmother, adept at keeping scammers on the line. If prompted to share bank details, the AI supplies fabricated numbers and names, leaving scammers empty-handed.
This AI operates with two goals: wasting scammers’ time and gathering insights into their tactics. By engaging scammers for extended periods - up to 40 minutes in some cases, Daisy ensures fewer real people fall victim. Additionally, the recorded interactions help refine O2's understanding of common fraud techniques.
The AI’s realism has elicited frustration from scammers, who often become increasingly irate as conversations drag on. Daisy’s development drew inspiration from popular YouTube scam-baiters, particularly Jim Browning, known for his skill in exposing and frustrating fraudsters.
“I Had an Onion on My Belt…”
Daisy operates by transcribing incoming calls into text, analysing the content via a custom language model layered with a character personality module, and generating responses in real time. A bespoke text-to-speech system then produces voice responses, making the interaction sound natural and lifelike.
A Tool for Protection, Not Retaliation
O2 hasn’t made Daisy accessible to customers for personal use. Instead, the AI interacts with scammers by answering calls directed to a list of numbers frequently targeted by fraudsters. Operating around the clock without human intervention, Daisy serves as a tireless deterrent.
Murray Mackenzie, Fraud Director at Virgin Media O2, highlighted the company’s broader efforts: “We’re committed to fighting scammers, investing in advanced technologies ranging from firewalls that block scam texts to AI-driven spam call detection. Daisy serves as a powerful reminder that the voice on the other end of the line isn’t always who they claim to be.”
Daisy’s creation stems from research indicating that 71% of Britons want to retaliate against scammers but hesitate due to the time commitment. This AI provides a solution, delivering a form of “revenge” without consuming users' time.
While Daisy’s effectiveness to fighting fraud with AI is impressive, the technology also raises concerns about its ability to simulate realistic conversations. Interestingly, scammers are increasingly deploying similar AI tools to impersonate trusted individuals, highlighting the dual-edged nature of this technology.
We are only just starting to scratch the surface of use-cases for AI. If you'd like to learn how AI can take over tasks in your business, give us a call at Digital Advantage - digitaladvantage.me
This is an incredible initiative by O2 to combat phone scams with technology. Daisy's ability to engage fraudsters while gathering insights into their tactics is a brilliant example of how AI can be used for social good. It’s fascinating to see AI deployed in such a clever and proactive manner. Looking forward to seeing how this tech evolves and continues to make a positive impact!