---
title: AI Data Handling and Providers
description: "Effective date: January 1, 2026 [Accept Mission](https://www.acceptmission.com/encyclopedia/accept-mission/) uses AI-assisted features by connecting our software to a pre-trained large language model (LLM) through Microsoft Azure AI Foundry in a private environment. Customer data is not used to train models. ## 1)…"
url: "https://www.acceptmission.com/knowledge/ai-data-handling-and-providers/"
updated: "2026-04-15"
category: "AI Transparency & Responsible AI"
---

# AI Data Handling and Providers

Effective date: January 1, 2026

[Accept Mission](https://www.acceptmission.com/encyclopedia/accept-mission/) uses AI-assisted features by connecting our software to a pre-trained large language model (LLM) through Microsoft Azure AI Foundry in a private environment. Customer data is not used to train models.

## 1) What AI features may process

AI features may process only what a user submits or selects for the task, such as:

- A user prompt (for example “summarize this” or “improve this text”)
- Selected workspace content the user chooses to include (for example an idea description or comments)
- Minimal task context (for example a requested structure or format)

We aim to minimize data sent for AI processing. Users should avoid including personal data unless their organization has approved that use.

## 2) What is stored

- Saved outputs: if a user saves an AI result into the platform, it becomes part of workspace content and is stored like other content.
- Prompts and temporary processing: prompts and temporary processing content are used to generate the result. They are not collected as a dataset for model training.

[Accept Mission](https://www.acceptmission.com/encyclopedia/accept-mission/) also maintains limited operational logs to run and secure the service. Access is restricted to authorized personnel.

## 3) Providers and training

We do not build or train our own LLM. AI requests are processed through Azure AI Foundry in a private environment. Providers do not use customer data, prompts, or outputs to train models.

## 4) Customer controls

Customers control AI usage through:

- Permissions and roles: who can access features in the workspace
- Governance: review and approval practices for how AI output is used
- Internal policies: what users are allowed to submit into AI features

## 5) How AI features are reviewed and approved

AI use is managed, not ad hoc. Before releasing or materially changing an AI feature, we review:

- Purpose and workflow value
- Data minimization and access control
- Privacy and security [Risk](https://www.acceptmission.com/encyclopedia/risk/)
- Human oversight and transparency (clear labeling and user review)
- Quality and safety checks

## 6) Report AI concerns

To report problematic AI behavior or concerns about data handling, email: security@acceptmission.com

Include the feature used, what you were trying to do, the input (remove personal data where possible), the output, and why it is a concern.