"Over 1,000 Scammers in a Year"... The True Identity of the Talkative 78-Year-Old Grandmother

O2 Develops 78-Year-Old Grandmother AI
‘Scambaiting’ Strategy Uses Virtual Number to Outsmart Scammers

In the United Kingdom, an artificial intelligence (AI) grandmother who confronts scammers has attracted attention as a new measure to reduce losses from voice phishing (telephone financial fraud).

On January 3, Japanese daily Asahi Shimbun and others reported that British telecommunications company Virgin Media O2 has developed an AI chatbot named 'Daisy,' designed as a 78-year-old grandmother, to prevent voice phishing crimes.

Daisy was created by applying state-of-the-art large language model (LLM) technology combined with voice synthesis. She can listen to the other party in real time and converse as naturally as a human. Since her release in November 2024, Daisy has been engaging scammers in lengthy phone calls, wasting as much of their time as possible.

AI Grandma 'Daisy' Fighting Scammers. Screenshot from O2 YouTube Channel

O2 secretly included Daisy’s phone number in the lists of potential scam victims that are illegally traded among voice phishing organizations. Unaware of this, criminals call her, mistaking Daisy for an easy target-an elderly person living alone. Whenever they ask for account numbers or personal information, Daisy deliberately tells unrelated stories or pretends not to understand, intentionally delaying the conversation.

Asahi Shimbun also introduced real examples of Daisy’s conversations. When a scammer tried to persuade her to install an app, Daisy suddenly shifted the topic to bread or knitting. If the increasingly frustrated scammer became annoyed, Daisy calmly responded, “You know, people talk more as they get older.”

According to O2, over the past year, more than 1,000 scammers have spoken with Daisy, and some wasted over 50 minutes on the call. This approach is known as 'scambaiting.' While scammers are occupied talking to the AI, there is less time for them to contact real victims.

However, O2 emphasized that while AI can waste scammers’ time, personal vigilance remains the most important factor in preventing voice phishing. “If you receive a suspicious call, the best thing to do is to hang up immediately and report it,” the company stressed.

Meanwhile, O2 was recognized for this achievement and received the 'Best Use of AI' award at the UK Marketing Awards last year.