LLMs can unmask pseudonymous users at scale with surprising accuracy

Ars Technica security feed covering emerging privacy, vulnerability, malware, supply-chain, phishing, cloud and infrastructure-security issues. Notable items include large-scale deanonymization using LLMs, Wi-Fi encryption bypass research, password-manager vault exposure, Lumma Stealer campaigns, Office exploitation by Russian state-linked actors, a Notepad++ update supply-chain compromise, exposed children’s AI chat transcripts, and destructive malware targeting Poland’s energy grid. The feed is a collection of reports rather than a single incident, and the provided metadata does not include-