Cybercrime trends show AI used for routine tasks amid rising banking scams

Recent research and federal investigations highlight the evolving landscape of cybercrime and the actual role of artificial intelligence in digital attacks. A study by researchers from Cambridge, Edinburgh, and Strathclyde suggests that AI has not yet triggered a massive hacking revolution. Instead, cybercriminals are using generative AI tools for routine tasks such as checking code, looking up answers, and avoiding manual reading. The study notes that safety guardrails on mainstream chatbots are largely effective, with jailbreak attempts typically failing within a week.

Separately, federal investigators have identified sophisticated scams involving fraudulent Google advertisements. Criminals have been using sponsored search engine links to direct banking customers to fake login pages, a scheme that has reportedly captured over 5,000 stolen credentials and caused millions of dollars in losses. The Justice Department recently announced the extradition of a Russian web developer accused of assisting in a large-scale bank account takeover operation using this method.