Meet Daisy: the AI granny driving phone scammers crazy
Virgin Media O2 built an AI bot named Daisy to troll phone scammers. Posing as a chatty 78-year-old grandmother who rambles about knitting and scone recipes, the bot kept fraudsters on the line for up to 40 minutes at a time to keep them away from real victims.

Phone scammers have a new worst nightmare, and her name is Daisy.
She is 78 years young, loves chatting about knitting patterns, and can never seem to find her reading glasses. She also happens to be an AI bot created by UK telecom giant Virgin Media O2 to do one thing: drive phone fraudsters absolutely crazy.
Scones, pastries, and 40 minutes of pure stalling
Instead of blocking fraudulent calls, O2 decided to waste the criminals' most valuable asset: their time.
Daisy was trained on hours of real scam audio. She knows precisely what keywords scammers listen for and exactly how to play dumb to keep them hooked. In recordings released by the carrier, one scammer claims Daisy's computer is riddled with viruses and demands remote access. Daisy responds by wandering off to look for her glasses, rambling about Internet Explorer, and confusing download instructions for Google Play with bakery pastries.
When one furious fraudster accused her of wasting his time, Daisy innocently replied that she only wanted a little chat. In another instance, she described her computer monitor going blank as "black like the night sky," while asking if the fraudster had any good local scone recommendations.
To bait the criminals, O2 teamed up with well-known YouTube scam-baiter Jim Browning. They seeded Daisy's phone numbers on shady "too good to be true" prize draw sites where scammers routinely scrape leads. Within days, the trap worked.
Daisy kept callers on the line for up to 40 minutes each. Because modern scam operations run like corporate call centers, frustrated fraudsters repeatedly passed her along when they failed to close the deal. In one case, Daisy was transferred between four different operatives before they gave up.
O2 marketing director Simon Valcarcel noted the initiative was rolled out briefly to raise public awareness, rather than as a permanent consumer tool. Some scammers eventually realized they were talking to an algorithm, though future iterations could feature varied accents and personas.
Why it matters
Scam prevention has traditionally been a game of whack-a-mole: carriers block a spoofed number, and fraudsters spin up ten new ones. Daisy flips the economic equation on its head. Scam call centers operate on volume and conversion rates. Tying up an operative for 40 minutes destroys their unit economics and mathematically prevents them from targeting a real victim during that window.
For consumers, however, you cannot yet outsource your annoying spam calls to Daisy. It remains a carrier-led PR and research experiment rather than a live feature on your smartphone. The arms race is also far from over. Once fraudsters start deploying real-time AI screeners of their own, we could soon witness the bizarre spectacle of two bots trapping each other in an infinite conversational loop.
Still, we think this is a masterclass in offensive defense. If network operators can eventually integrate conversational honeypots directly into the telecom infrastructure, scammers will no longer know if their target is an easy mark or an endless conversation about scones.
Sources (1)
- theguardian.com(opens in a new tab)Originaltheguardian.com
Topic trackerO2's AI scam-baiting bot Daisy
- Meet Daisy: the AI granny driving phone scammers crazyThis story
- O2's new AI granny wastes scammers' time with endless rambling