AI technology is increasingly being utilized to combat cybercrime, as governments face challenges in addressing online criminal activities. One notable example is the Australian company Apate, which has developed a system that diverts phone scammers to AI bots designed to prolong conversations without falling for scams. Dali Kaafar, the founder and CEO of Apate, stated that this approach aims to keep scammers engaged, thereby preventing them from reaching potential victims. Apate's platform reportedly utilizes around 350,000 bots that not only handle phone calls but also infiltrate scam chat groups and respond to messages. The company claims to have gathered over 250,000 pieces of information related to fraudsters, including scam URLs and bank details.
The bots are designed with various personalities and language skills to avoid detection by scammers. Kaafar noted that conversations with scammers can last over two hours, and during testing, the AI system demonstrated natural conversational timing. Despite these efforts, the prevalence of scam calls and texts remains high, with cybercriminals generating billions of messages annually.
Experts suggest that AI could enhance anti-scam initiatives by improving intelligence sharing among law enforcement, social media companies, and financial institutions. Additionally, innovative strategies are being explored to exploit the psychological vulnerabilities of cybercriminals. For instance, honeypots, which are decoy systems used to attract hackers, have begun incorporating large language models (LLMs) to increase their effectiveness. Research from ETH Zurich indicates that LLM-powered honeypots can keep attackers engaged longer than traditional models, providing valuable insights into hacking techniques.