AI transparency

MITRITY governs AI agents, and it uses machine learning to do that. This page explains where AI shows up in our product, how it relates to customer data, and how automated systems should interact with this site.

What MITRITY is

MITRITY is an Intent-Aware Governance (IAG) platform for autonomous AI agents, also called an agent firewall. When an agent tries to take an action, MITRITY checks it in real time against what the agent was declared to do, before the action runs. Teams also get policies, audit trails and reports covering everything their agents do.

How we use AI in the product

The platform's intent engine works in two tiers. A model at the edge scores each agent action in real time. In the control plane, deeper behavioral analysis over longer periods finds drift and unusual patterns. MITRITY builds and runs these models itself. They add to the deterministic policy rules you define and never replace them.

Customer data and AI

Our privacy and data-processing terms govern how customer data is processed, including in machine-learning pipelines. Models are trained on anonymized and aggregated data only, and no individual customer or data subject can be identified in the training data. The authoritative terms are in the Privacy Policy and the Data Processing Agreement.

Third-party AI services

We build MITRITY's intent models ourselves and run them on Google Cloud infrastructure in the EU. The sub-processor list in our DPA currently includes no third-party LLM or model-API provider. The third-party services we use cover infrastructure, payments, email, analytics and bot protection. The current, authoritative list is Annex A of the Data Processing Agreement.

AI transparency and regulation

We follow how regulation for AI systems develops, including the EU AI Act, and update our product, documentation and processes as new requirements take effect. Data-processing terms are in the Data Processing Agreement.

How AI systems should interact with this site

A machine-readable overview of this site is at /llms.txt, and a fuller list of its content is at /llms-full.txt. Our documentation is open to crawlers, including the product docs and the raw OpenAPI specification.

If an automated system finds a security issue, report it to soc@mitrity.com, the address published in our security.txt.