Don’t expect quick fixes in ‘red-teaming’ of AI models. Security was an afterthought

Current AI models are simply too unwieldy, brittle and malleable, academic and corporate research shows. Security was an afterthought in their training as data scientists amassed breathtakingly complex collections of images and text. They are prone to racial and cultural biases, and easily manipulated. — AP

BOSTON: White House officials concerned by AI chatbots’ potential for societal harm and the Silicon Valley powerhouses rushing them to market are heavily invested in a three-day competition that ended Sunday (Aug 13) at the DefCon hacker convention in Las Vegas.

Some 2,200 competitors tapped on laptops seeking to expose flaws in eight leading large-language models representative of technology’s next big thing. But don’t expect quick results from this first-ever independent “red-teaming” of multiple models.