LinkedIn Phishing Scam Uses Fake Notifications to Hijack Accounts

Multiple high-impact threats reported: a LinkedIn phishing campaign using fake notifications and lookalike domains to steal credentials and hijack professional accounts; a human-error data leak at Anthropic exposing ~512,000 lines of Claude AI code and internal secrets; ransomware actors abusing legitimate IT/admin tools (the "dual-use" dilemma) to bypass AV; a critical ImageMagick zero-day enabling remote code execution via image uploads affecting Linux distributions and WordPress deployments; a high-profile npm supply-chain compromise of the axios package; and an actively-exploited, upgraded