O2, the UK’s largest mobile network operator, has recently introduced a clever chatbot named “dAIsy” aimed at frustrating phone scammers. This chatbot, designed to mimic an older woman with a penchant for discussing topics like knitting and her cat Fluffy, is intended to keep fraudsters engaged in meaningless conversation, preventing them from obtaining any sensitive information.
According to a press release from O2, the “AI Granny” utilizes a combination of various AI models to transcribe the caller’s voice into text, generate a response using a custom large language model, and then convert it into a voice answer using a text-to-speech model. Notably, the AI was trained with the assistance of Jim Browning, a renowned “scambait” expert with a substantial YouTube following.
The effectiveness of dAIsy can be witnessed in action, as demonstrated in a video provided by O2. If successful, this innovative approach could potentially make a significant impact in combating telephone scams. Recent reports from the FBI indicate that individuals over the age of 60 fell victim to telephone scams, resulting in losses amounting to $3.4 billion in 2021, up from $3.1 billion in the previous year. With the rise of generative AI and voice impersonation technology, these numbers are expected to rise even further in the future.
As seen in the video demonstration, dAIsy effectively engages with scammers, wasting their time and preventing them from carrying out their malicious activities. This proactive approach by O2 highlights the potential of AI in combating fraudulent schemes and protecting vulnerable individuals from falling victim to scams. By leveraging cutting-edge technology and innovative strategies, companies like O2 are taking important steps towards safeguarding consumers and enhancing cybersecurity measures in an increasingly digital world.
