Meet Daisy — the AI-generated granny helping to trap scammers
Scammers are using increasingly sophisticated methods, including artificial intelligence and deepfakes, to deceive their victims. However, the tables are turning as AI is now being used to combat these fraudulent activities. Telecom company O2 has deployed an AI-powered tool named Daisy, designed to engage with scammers and waste their time.
Daisy, with the persona of a grandmother, has proven to be effective in keeping scammers on the phone for extended periods. According to O2, 67% of British individuals are concerned about falling prey to fraud, making Daisy a valuable asset in the fight against scammers.
The Creation of Daisy
Daisy was created by the team at O2 using a custom-trained large language model with a 'character personality layer' to provide personalized responses. The AI tool listens to the caller, generates responses, and engages in conversations to deter scammers.
Similar to Google Gemini Live and Meta AI Voice, Daisy interacts with scammers in real time, keeping them engaged in lifelike conversations. The goal is to prolong the interaction without any human intervention, frustrating scammers with fake information and details.
Effectiveness of Daisy
A video demonstration by O2 showcases Daisy's ability to outsmart scammers, causing them to become frustrated with the conversations. By maintaining the scammer on the line, Daisy helps prevent individuals from falling victim to fraudulent activities.
Murray Mackenzie, Director of Fraud at Virgin Media O2, commended Daisy for outsmarting scammers and recommended forwarding suspicious calls to 7726 for investigation. The use of AI tools like Daisy is essential in combating fraud and protecting individuals from potential scams.
With Daisy's success in deterring scammers, individuals can feel more secure in their interactions and be proactive in preventing fraud. The innovative use of AI technology in trapping scammers marks a significant advancement in consumer protection.
For more information on AI technology and fraud prevention, visit Tom's Guide AI section.
