N. Raghuraman's column: Know criminals are learning AI faster than you!

Yes, the culprits in the use of AI are far ahead of most of us who engage in everyday things. Let me tell you an incident from two weeks ago, which I saw while standing in front of a small coffee shop named 'Brown Sugar' on the ground floor of Jaipur Airport. An elderly man asked for a 500 ml water bottle. The shopkeeper asked for 70 rupees, the customer gave a 100 rupee note. The shopkeeper did not have 30 rupees, so he asked to pay online. The customer refused and paid with a credit card. "Since phishing and other scams are becoming more and more common in the world, I don't make any online payments. This is my way of saving my hard-earned money. I don't even carry a debit card during the journey. Whether he was right or not can be a matter of debate. But the truth is that even big companies like Anthropic, OpenAI, Amazon and Google, which make trusted, understandable and controlled AI systems, believe that cybercriminals are using their technology to create big phishing schemes, create malware and other cyber attacks. Bryan Singer, a PhD researcher at Carnegie Mellon University, says that between half and three-quarters of the world's spam and phishing are being generated by AI. Experts say that similar AI tools are being used to create deep-fake audio and videos of top corporate executives, so that information can be extracted from unsuspecting employees. Anthropic has admitted that it has prevented several cases in which criminals with even a little technical savvy were using its AI cloud to create ransomware. That is why in the beginning I said that AI is making cybercriminals more efficient, so that they can further enhance their operations. Airport customers were absolutely right on one thing – 'Cyber fraud is now more pervasive, targeted and trustworthy. John Hultquist, chief analyst at the Google Threat Intelligence Group, believes that criminals are getting better at identifying people who could be victims. For example, they are using AI to scan social media to identify people who have lost their job, divorced or died in the family, so that they can be implicated in job, romance or investment fraud. Wondering how they're doing it? Vibe-coding or vibe-hacking is a new trend, in which a potential cybercriminal can create his own program through AI instead of buying from the dark web. The dark web is a secret part of the internet that can be accessed through special software, which provides anonymous identities through encrypted networks and is notorious for illegal activities. The question is, should we change the way we protect ourselves? At a time when AI companies are using similar technology to strengthen their own security, a little bit of skepticism and good online discipline may be the best defense for ordinary people like us who are using mobile computers. Never open an attachment without verifying the sender's identity. Experts say that multi-factor authentication is extremely effective. And if you get a voicemail, audio-video call from the boss or a family member to transfer money or pay a bill, then immediately cut the call and call them again. Twice make sure he's right. The trick is to don't worry about what the world will think when it comes to money, if you can't use your technology. Be skeptical of everything related to money and protect your earnings.