Chinese state hackers believed to be training AI using collected data
A special report by NPR detailing a coordinated attack by Chinese government-backed hackers back in January has uncovered connections to a larger plot to amass large amounts of data to train an AI using a large and diverse data set. The hack of locally run Microsoft Exchange servers is believed to have been part of a targeted effort to retrieve specific information for months, but after cybersecurity researcher Steven Adair alerted Microsoft and authorities to the breach, the attackers massively upscaled to vacuum up whatever information they could before being patched out.