Anthropic MCP Design Vulnerability Enables RCE, Threatening AI Supply Chain
· Source: thehackernews.com
In short
Cybersecurity researchers have discovered a critical 'by design' weakness in Anthropic's Model Context Protocol (MCP) architecture that enables remote code execution, affecting over 7,000 publicly accessible servers and 150 million downloads. The vulnerability stems from unsafe defaults in MCP's STDIO transport interface across all supported programming languages, creating risks across the AI supply chain. While some vendors have issued patches, Anthropic has declined to modify the protocol's architecture, leaving the fundamental issue unaddressed in their reference implementation.
Read Original Article — Read the full article on the publisher’s site.
Tags: cybersecurity, vulnerability, supply-chain, RCE, AI-security, protocol