AI Transparency Policy

Two places inside the product:

• Participant reflection. After a participant answers the three weekly questions, an AI model generates a short reflection back to them: a mirror, a question to sit with, and something to notice this week.
• Leader weekly summary. Once the week is closed, an AI model summarizes the group's responses into a one-page pattern view for the leader.

AI does not evaluate individual people, assign scores, produce risk ratings, make diagnoses, or take actions on behalf of a leader or organization. It does not monitor Tap In in real time.

The weekly summary is a starting point for a leader's own thinking and conversations. Any decision about how to support a person is made by that leader, their team, and their organization — never by the model, and never by Tap In on their behalf.

Responses are sent to our AI model provider solely to generate the reflection and summary, under terms that prohibit use of that data to train their models. Tap In does not train its own models on participant responses.

AI-generated text can be wrong, incomplete, or miss context. Reflections and summaries should be read as a starting point for human attention, not as a verdict.

If something in a reflection or summary feels off, trust the person in front of you over the words on the page.

We use established AI model providers accessed through our platform gateway. The specific model may change over time as better models become available; changes are made without altering the promise on this page.

See our Privacy Policy and Safety Policy for more on how data is handled and how Tap In fits alongside real support.