2023-06-27 15:22:36
The scammers will be caught at their own game! Researchers have developed Apate, an AI-powered chatbot capable of speaking out loud in natural language. Its goal: to keep scammers online as long as possible. Well seen !
Scams are a real pain, especially when they are telephone, because you have to manage to get out of the scammers – unlike SMS and emails, which you just have to delete in two seconds – and that makes us lose a precious time. All it takes is for our phone number to end up on the Dark Web, following a successful hack or phishing, for us to find ourselves harassed. GDPR or Banque de France scams, canvassing for the personal training account (CPF), vishing… Cybercriminals are not lacking in imagination to manipulate us!
Phone scams are an extremely lucrative business that grows every year. In 2022, l’ACCC (Australian Competition & Consumer Commission) estimates that Australians have lost over €3.1 billion to scammers. Because even if telephone operators today block millions of fraudulent calls, users are still inundated with scams. And it’s not going to get any better… What if, to prevent scammers from scamming, you had to catch them at their own game? This is the original idea, to say the least, that researchers from Macquarie University in Australia had. Under the leadership of the professor and executive director of the establishment’s cybersecurity center Dali Kaafarthey decided to waste their time on scammers by keeping them online with a chatbot, dubbed Get it – in reference to the Greek goddess of deceit, perfidy, fraud, deception, deceit and dishonesty.
Apate: an AI to waste the time of crooks
The idea for such a chatbot came to Dali Kaafar when he received a call from a scammer during a dinner with his family. Instead of hanging up, he decided to chat with him in order to make his children laugh. Result: they stayed 40 minutes on the phone. 40 minutes to say the least time-consuming during which the scammer could not make victims. “I realized that while I had wasted the scammer’s time so he couldn’t reach the vulnerable, which was the point – it was also 40 minutes of my own life that I didn’t wouldn’t recover”explains Professor Kaafar on Macquarie University website. “Then I started thinking about how we could automate the whole process and use natural language processing to develop a computerized chatbot that could have a believable conversation with the scammer.” This is how Apate was born. “We are very excited that this new technology can break the business model of fraudulent calls and make it unprofitable”says the researcher.
To develop Apate, the Macquarie University Cybersecurity Center team began by analyzing scam phone calls and identifying the social engineering techniques used by scammers on their victims – creating a sense of urgency, embodying an authority figure, threatening punishment, etc. – using machine learning techniques and natural language processing to identify typical scam scenarios. For this, the researchers used a dataset of real-world fraudulent conversations, including call recordings, email transcripts and social media chats, so that the chatbot could generate its own conversations in a scam situation. Professor Kaafar says he was surprised by the adaptability of Apate, who would have reacted well to situations in which she had not been trained.
The researcher explains that, thanks to advances in the field of natural language processing (NLP) and human voice cloning by artificial intelligence, it is now possible to create AI-powered conversational agents that are able speak in natural language, adopt a specific persona, and maintain a compelling conversation through consistent responses. “The chatbots we’ve developed can trick scammers into thinking they’re talking to viable scam victims, so they spend time trying to scam the bots”, explains Professor Kaafar. The technology, which has already been the subject of a patent application, is still under development. For now, Apate only occupies scammers 5 minutes, instead of the targeted 40 minutes. It could be of interest to operators – the researchers are also in discussion with some of them –, but also various institutions (governments, banks, etc.) by collecting a lot of data relating to the evolution of scams.
1687880343
#chatbot #scam #scammers