Safety consultants flag one other worrying problem with Anthropic AI techniques – this is what they discovered

Anthropic’s MCP Inspector mission carried a flaw that allowed miscreants to steal delicate information, drop malwareTo abuse it, hackers have to chain it with a decades-old browser bugThe flaw was mounted in mid-June 2025, however customers ought to nonetheless be on their guard

The Anthropic Mannequin Context Protocol (MCP) Inspector mission carried a critical-severity vulnerability which might have allowed menace actors to mount distant code execution (RCE) assaults towards host gadgets, consultants have warned.