---
title: "AI Use"
description: "How Chudaxi uses AI in its own work and in the systems it builds for clients."
layout: legal
legal_review: pending
updated: 2026-10-01
---

Chudaxi is an AI-native company. This page explains where AI is involved and what we do to keep it accountable.

## In our own work {#our-work}

AI agents help us research, write, develop software and run day-to-day operations, including this website. A person reviews what we publish and what we ship. When we state facts about our work, they come from records, not from a model.

## In what we build {#what-we-build}

- Systems we build may use third-party AI models or open-weight models hosted for the client. We tell clients which models a system uses.
- **We do not use client data to train models**, and we choose providers and settings under which submitted data is not used for training.
- AI agents that talk to the public identify themselves as automated when asked, and offer a route to a person.
- Agents work with the permissions a task needs. Actions with significant effects are logged and, where they matter, approved by a person.

## Limits {#limits}

AI output can be wrong. Our systems are designed so that important answers are grounded in the client's own data, and uncertain cases are handed to a person. Product visuals marked "Illustrative" on this website do not show real client data.

## Contact {#contact}

Questions about how we use AI: kris@chudaxi.com.