Honest about the machine.

1. What Tegra is

Tegra is Integro's reflective companion. It reads the context you give it — your journals, recent practice, stated intentions — and offers prompts, summaries, and observations. It is built on large language models provided by Anthropic, and is tuned by Integro to stay within the scope of integration work.

When you use any AI feature within the Platform, you are interacting with an artificial intelligence at all times — never with a licensed therapist, counselor, psychiatrist, or medical professional — unless the interface explicitly states otherwise.

2. What Tegra is not

Tegra is not a therapist, doctor, coach, or crisis counselor. It cannot diagnose a condition, recommend a medication, assess your clinical state or safety risk, contact your care team or emergency services on your behalf, or provide emergency care.

Tegra does not have persistent memory of prior sessions unless you have explicitly enabled that feature — and even where enabled, that memory is limited in scope. Tegra may not always recognize when your needs exceed what it can appropriately support; you are responsible for making that judgment and seeking appropriate help.

If you are in crisis, please contact 988 (US) or your local emergency service — Tegra will direct you there too.

3. How AI is used

Our AI systems analyze information you provide, along with other relevant context, to assist with responses, recommendations, and guided reflection. AI may influence:

• The responses and recommendations our AI agents provide to you
• How our AI agents evaluate what information is relevant to your session
• Eligibility for certain groups or in-app features

Right now, no AI system in the Integro app is being used to make consequential decisions about you — that is, decisions with a material legal or similarly significant effect on your access to healthcare, employment, housing, financial, or government services.

4. Human oversight

Our AI systems do not operate on their own. Human review, oversight, and safeguards are in place to help ensure responses are fair, accurate, and appropriate.

A small internal team periodically reviews sampled transcripts to catch safety issues and improve Tegra's behavior. Any review is governed by strict access controls and confidentiality obligations. You can opt out of sample review in Me → Data & Privacy → Privacy Settings without losing access to Tegra.

5. How your data is used

When you chat with Tegra, your messages are sent to Anthropic's API along with the context we supply. Anthropic does not train its models on content sent through the API on paid tiers, and we do not share your journals with any third party for model training — ours or otherwise.

Transcripts are stored in your Integro account so you can return to them. You can delete any conversation at any time from within the app.

6. Where models can be wrong

Language models sometimes generate confident-sounding text that is inaccurate, incomplete, or mismatched to your situation. Responses are generated based on patterns in language data — they may occasionally be imprecise, contextually limited, or insufficient for complex psychological or medical needs. We label AI-generated content clearly and ask you to treat it as a starting point for your own reflection — not as fact.

Trust your own experience and the people who know you. Use Integro as a complement to your clinical care, not a replacement for it.

7. Your rights regarding AI

You have the right to:

• Request information about how an AI-assisted decision that affects you was made
• Correct inaccurate personal data that informs AI outputs
• Appeal or request human review of certain decisions, where applicable

We take reasonable steps to identify, assess, and reduce the risk of algorithmic discrimination and other potential harms. To exercise any of these rights, contact us at legal@integro.today.

8. Changes to how Tegra works

We will update this page whenever we change models, add new AI features, or adjust how data flows through them. Material changes will be noted in-app before they take effect.

Questions about this document? Reach us at legal@integro.today.