The 'AI confidence trap' many New Zealanders are caught in (www.rnz.co.nz)

Recent research from Avast highlights an alarming trend in New Zealand known as the "AI confidence trap," where 66% of adults believe they can identify AI-generated fakes, yet many still share sensitive information with these tools. The Avast 2026 Safe Tech Report reveals that 40% of respondents have considered sharing personal documents like CVs and work-related files with AI, despite knowing the risks. This paradox suggests that convenience often outweighs caution when it comes to interacting with AI applications, which can imperil personal data security. Stephen Kho from Avast warns that the information people provide to AI tools may not only be used for model training but could also fall into insecure channels or be exploited in scams. He cites a recent incident involving deep fake technology where a victim was manipulated into transferring $20 million after being deceived by what seemed like a legitimate video call. As AI makes creating scams more sophisticated and harder to detect, Kho emphasizes that users should treat AI chatbots with the same caution they would apply to sharing information with someone in person, reminding them that these tools lack the legal protections of traditional professionals.