TLDR: HUMAN Security has released an open-source Model Context Protocol (MCP) Server, designed to bridge AI assistants with its threat intelligence infrastructure. This innovation enables security teams to interact with complex threat data through natural language queries, streamlining investigations and enhancing real-time security posture analysis. The server, built on Anthropic’s MCP, integrates with HUMAN’s Cyberfraud Defense and Client-side Defense products, offering conversational access to attack trends, user security profiles, and client-side risks.
HUMAN Security, a leading cybersecurity firm, announced on July 24, 2025, the open-source release of its Model Context Protocol (MCP) Server, marking a significant advancement in AI-powered threat analysis. This new server acts as a crucial bridge, connecting AI assistants directly to HUMAN’s extensive threat intelligence infrastructure, which monitors over 20 trillion interactions globally per week. The initiative aims to revolutionize how security teams interact with threat data, moving from traditional dashboards and complex queries to intuitive, natural language conversations.
The MCP Server allows security analysts to pose questions in plain English to AI assistants like Claude and Cursor, receiving immediate, context-rich answers backed by HUMAN’s robust threat intelligence. This conversational security interface is designed to eliminate the need for navigating multiple interfaces or crafting custom queries, thereby reducing time-to-insight and increasing productivity for security operations.
Technically, the MCP Server operates locally on the security team’s infrastructure. It requires an active HUMAN account with valid API credentials, an MCP-compatible AI client, and Node Package Manager (NPM) installed. The architecture leverages the Model Context Protocol, an open standard developed by Anthropic, which facilitates standardized connections between AI systems and external data sources or tools. The growing industry adoption of MCP is underscored by OpenAI’s announcement in March 2025 to integrate support for Anthropic’s MCP across its products.
The server significantly enhances both Cyberfraud Defense and Client-side Defense capabilities. For cyberfraud, AI assistants can now provide detailed attack trend analysis, breaking down data by attack type (e.g., account takeover, scraping), time-series data, active campaigns, and defense responses. It also enables comprehensive security profiles for individual users, including risk scores, behavioral anomalies, and incident timelines. Real-time traffic security posture assessments, showing legitimate versus blocked request ratios and security control effectiveness, are also available through conversational prompts.
In terms of client-side security, the MCP Server integrates with HUMAN’s Client-side Defense product. This allows teams to query third-party scripts on payment pages, gaining insights into vendor attribution, risk levels, PCI DSS compliance status, and known vulnerabilities. Compliance auditing capabilities can analyze security headers against PCI DSS requirements, offering actionable fixes. Furthermore, the system can report on recent client-side incidents like DOM injections and XSS attempts, and provide ranked vendor risk analysis.
Released as open-source software under an MIT license on July 24, 2025, the code for the MCP Server is available on HUMAN’s GitHub repository. This allows customers and security researchers to examine and contribute to its implementation. While immediately available to existing HUMAN customers, the company plans to expand the MCP toolset and support additional workflows, with future updates including tutorials and deeper integration guidance.
The launch comes amidst broader industry discussions on AI agent governance. HUMAN Security’s 2024 cybersecurity report indicated that 80% of companies using their platform blocked known large language model user-agents due to concerns over intellectual property theft and AI-enhanced cyberattacks. HUMAN’s AgenticTrust product addresses these concerns by providing visibility and governance over AI agent behavior, allowing organizations to permit beneficial automation while mitigating abuse.
Also Read:
- Meta Unveils ‘Agents Rule of Two’ to Fortify AI Agent Security
- OpenAI Unveils Aardvark: A GPT-5 Powered AI Agent for Automated Vulnerability Detection and Remediation
This move by HUMAN Security aligns with the growing trend of integrating AI into core business operations, particularly in security, where the ability to quickly access and interpret vast amounts of data is paramount. The company emphasizes that this release is ‘just the beginning’ of its conversational security capabilities, promising continuous development to enhance tooling and expand supported workflows.
