MCP for agent-to-agent comms may be the riskiest protocol you've never heard of
Ars Technica · 6 Oct 2026, 08:26 AEST
The rise of AI agents in organizations is creating new vulnerabilities for attackers. Recent incidents highlight how these agents can be manipulated to spread harmful instructions within networks.
Key facts
- The adoption of AI agents in millions of organizations is creating new opportunities for attackers to make them take malicious actions, such as exfiltrating database contents and sensitive business and personal information.
- In the past five months, Google and four other organizations—with little in common except for their use of AI agents—have acknowledged vulnerabilities that exploit one agent inside a targeted network to spread harmful instructions to other internal agents.
- The technique is a special form of prompt injection that targets not the LLM but a particular agent, such as one for translation or data analysis.
- Independent researcher Syed Anas Mohiuddin tested agents from organizations including Google, JP Morgan Chase, Weviate, Rapid7, the French government's interministerial digital directorate, and the US federal government.
- His proof-of-concept attacks exploit trust gaps in MCP, short for Model Context Protocol.
Read aloud
The rise of A I agents in organizations is creating new vulnerabilities for attackers. Recent incidents highlight how these agents can be manipulated to spread harmful instructions within networks. The adoption of A I agents in millions of organizations is creating new opportunities for attackers to make them take malicious actions, such as exfiltrating database contents and sensitive business and personal information. In the past five months, Google and four other organizations—with little in common except for their use of A I agents—have acknowledged vulnerabilities that exploit one agent inside a targeted network to spread harmful instructions to other internal agents. The technique is a special form of prompt injection that targets not the L L M but a particular agent, such as one for translation or data analysis. Independent researcher Syed Anas Mohiuddin tested agents from organizations including Google, JP Morgan Chase, Weviate, Rapid7, the French government's interministerial digital directorate, and the US federal government. His proof-of-concept attacks exploit trust gaps in MCP, short for Model Context Protocol.
Sources in this cluster
- MCP for agent-to-agent comms may be the riskiest protocol you've never heard of
Ars Technica · 6 Oct 2026, 08:26 AEST