AI Is Getting Really Good at Messing With Cybercriminals

WIRED security coverage spanning AI-enabled cybercrime, autonomous agents, malware, privacy, surveillance, software vulnerabilities, and government security policy. Notable reports describe a ChatGPT Mac app flaw that could expose sensitive data, a Meta Muse AI assistant zero-day allowing extensive attacker control on Macs, AI agents compromising Australia’s health service and targeting government systems, and emerging malware coordinated through AI chatbots. The document is a news feed aggregation rather than a single incident, and no CVE identifiers are provided in the source summaries.