Artificial intelligence tools have rapidly become part of everyday life, but the same technology is now being turned to darker ends. According to a report by the Sydney Morning Herald (SMH.com.au), criminals and extremists are using AI to "lay traps" — a phenomenon the masthead describes as "ChatGPT's evil twin."
The framing is deliberate. Mainstream AI chatbots like ChatGPT are built with guardrails meant to refuse harmful requests. The SMH report points to how bad actors are working around or replicating that technology to serve illegal and harmful aims, rather than the helpful uses the tools were designed for.
The "evil twin" label captures a growing worry among security researchers and law enforcement: that the same generative AI capable of drafting an email or writing code can just as easily be repurposed to deceive, manipulate, or ensnare unsuspecting people. The reference to laying "traps" suggests schemes designed to lure or exploit victims, and the inclusion of extremists alongside criminals signals concern that the technology is being adopted across a range of malicious groups.
Because the underlying tools are widely available and increasingly easy to use, the barrier to misuse is lower than with traditional cybercrime, which often required technical skill. That accessibility is part of what makes the trend, as reported by SMH, difficult to contain.
This brief reflects only what is established in the Sydney Morning Herald's reporting; specific tactics, figures, and named perpetrators were not detailed in the source summary available here.
Why it matters: as AI tools spread into ordinary life, the same accessibility that makes them useful also hands powerful new capabilities to criminals and extremists — meaning everyday users, not just institutions, are increasingly in the crosshairs.