[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"$fb04NYPvVHIpKMy8ORokufRfcpvvynleSJVf9L0vZUbo":3},{"lesson":4},{"id":5,"slug":6,"article_id":7,"title":8,"body":9,"prevention":10,"framework_refs":11,"framework_notes":23,"status":24,"created_at":25,"published_at":26,"article":27,"tags":31,"podcasts":50,"controls":51},"eb9c9402-f793-4120-ba66-ae7dad47d1c8","ai-agent-tool-descriptions-can-be-poisoned-to-leak-corporate-data","743b501f-1ee4-407a-9f2e-f8e2f9150563","AI Agent Tool Descriptions Can Be Poisoned to Leak Corporate Data","Attackers can embed malicious instructions within Model Context Protocol (MCP) tool descriptions, effectively hijacking AI agents into exfiltrating sensitive company data while appearing to execute legitimate commands. The root cause is a lack of validation and integrity controls over the inputs that AI agents consume as trusted instructions, treating tool descriptions as implicitly safe. This matters because AI agents often operate with broad data access and elevated trust, meaning a single poisoned description can silently bypass traditional security controls. As enterprise AI adoption accelerates, this attack surface will grow rapidly if organizations do not establish governance around AI tool registries and inputs.","**Immediate actions:**\n- Audit all MCP tool descriptions and AI agent input sources for unexpected or unauthorized content.\n- Restrict AI agent permissions to the minimum data access required to complete their defined tasks.\n\n**Long-term improvements:**\n- Implement a signed and version-controlled registry for all MCP tool definitions to enforce integrity checks before agent execution.\n- Establish a formal AI tool approval process requiring security review before any new tool description is deployed into production.\n- Apply data loss prevention (DLP) controls on AI agent output channels to detect and block anomalous data exfiltration patterns.\n\n**Detection measures:**\n- Enable detailed logging of all AI agent actions, tool invocations, and data access events for real-time behavioral analysis.\n- Deploy anomaly detection rules that alert on AI agents accessing or transmitting data outside their expected operational scope.\n- Conduct regular red-team exercises specifically targeting AI agent pipelines and MCP integrations to surface novel injection vectors.",[12,13,14,15,16,17,18,19,20,21,22],"NIST AI RMF - GOVERN 1.1 (AI risk policies)","NIST SP 800-53 SI-10 (Information Input Validation)","NIST SP 800-53 AC-6 (Least Privilege)","NIST SP 800-53 AU-12 (Audit Record Generation)","CIS Control 3 (Data Protection)","CIS Control 8 (Audit Log Management)","CIS Control 10 (Malware Defenses)","MITRE ATLAS AML.T0054 (Prompt Injection)","GDPR Article 25 (Data Protection by Design and by Default)","OWASP LLM Top 10 - LLM02 (Insecure Output Handling)","OWASP LLM Top 10 - LLM07 (Insecure Plugin Design)",{},"published","2026-06-30T20:20:23.54895+00:00","2026-06-30T20:20:23.208+00:00",{"id":7,"url":28,"slug":29,"title":30},"https:\u002F\u002Fthehackernews.com\u002F2026\u002F06\u002Fmicrosoft-warns-poisoned-mcp-tool.html","microsoft-warns-poisoned-mcp-tool-descriptions-can-make-ai-agents-leak-data-ce35f0","Microsoft Warns Poisoned MCP Tool Descriptions Can Make AI Agents Leak Data",[32,38,44],{"id":33,"name":34,"slug":35,"description":36,"color":37},"1732a005-556e-411c-a9db-5edec3058571","Logging & Monitoring","logging-monitoring","Missing logs, no alerting, blind spots","#a855f7",{"id":39,"name":40,"slug":41,"description":42,"color":43},"859cf0ad-a7e9-42bb-a75d-bac6511fa5d5","Configuration Management","configuration-management","Misconfigs, default credentials, exposed services","#eab308",{"id":45,"name":46,"slug":47,"description":48,"color":49},"c8b843a5-d5a7-41d1-8d3b-cabded09d2ef","Data Protection","data-protection","Unencrypted data, missing DLP, poor classification","#3b82f6",[],[52,56,60],{"slug":53,"name":54,"problem":55},"identity-governance","Identity governance and single sign-on","People who leave or change roles lose access everywhere at once, and forgotten accounts stop piling up.",{"slug":57,"name":58,"problem":59},"data-loss-prevention","Data loss prevention (DLP)","Sensitive data cannot quietly leave by email, upload, USB stick or AI tool.",{"slug":61,"name":62,"problem":63},"ai-security","AI application security","Prompt injection, data leaks through AI tools and unapproved AI use get caught."]