TechThe Register3h ago
Anthropic won't own MCP 'design flaw' putting 200K servers at risk, researcher says
TL;DRSecurity flaw in Anthropic's MCP protocol could compromise 200,000 servers.
Why it matters: Widespread vulnerability in AI infrastructure tool threatens enterprise deployments and server security.
Bug or feature? A design flaw – or expected behavior based on a bad design choice, depending on who is telling the story – baked into Anthropic's official Model Context Protocol (MCP) puts as many as 200,000 servers at risk of complete takeover, according to security…
Read full articleSource: The Register · Opens in new tab