Security vulnerabilities currently being leveraged by threat actors in the wild, along with available patches and mitigation strategies.
Follow this story through dated reporting and the underlying records. Evidence labels distinguish documented findings, attributed reporting, disputed accounts and claims that remain unverified. Start with the background reports below, then read the latest developments.
Anthropic reveals a Claude research model developed a zero-click exploit for a critical Linux kernel flaw, while hackers chain Ivanti zero-days to steal credentials.
