Help Centre/Data & Privacy/AI Training and Data Use

AI Training and Data Use

Category: Data & Privacy Last updated: March 2026 Reading time: ~3 min

Overview

A key concern for practitioners using any AI-powered platform is whether their data — or their clients' data — is used to train AI models. The short answer for TheraScripts: no, it is not.

Our AI processing arrangement

TheraScripts uses an AI processing partner whose API operates under a zero-data-retention policy for our account. This means:

  • Prompts sent to the API are not retained beyond the duration of the API call
  • Outputs returned are not stored by the AI provider
  • None of our account's data is used to train or improve the underlying models

This is a specific commercial API arrangement, distinct from how consumer-facing AI products work. Consumer products may retain and use inputs for model improvement. Our API arrangement explicitly prohibits this.

If our AI processing partner changes at any point, we will update this article and notify users. Any replacement provider will be required to meet equivalent or stronger data handling standards as a condition of use.

What is sent to the AI

When generating a script, a structured session snapshot is sent to the API. This snapshot contains:

  • Session goal and structure parameters
  • Client profile data in non-identifying format (no names, no contact details, no PII)
  • Therapeutic style and delivery mode configuration
  • Module selection context

Nothing in the prompt identifies an individual client. The prompt is assembled from structured fields — not from free-text clinical notes.

What TheraScripts stores

The generated script output is stored in your workspace so you can access, edit, and export it. This is retained according to our standard retention policy. It is not shared with AI providers or any third party.

KYODO assistant

The KYODO clinical knowledge assistant uses retrieval-augmented generation, drawing on a curated clinical knowledge base. The same zero-retention principle applies: conversation inputs are not retained by the AI provider and are not used for training.

How to verify this

You can request a copy of our Data Processing Agreement, which details the sub-processors we use and the contractual obligations they operate under, including data retention and training restrictions. Contact us via the contact page to request a copy.

Related articles