AI Disclosure & Limitations Statement

This statement explains, in plain language, how AI works on LifeChoices and what its limits are. We're publishing it as its own document — not buried in the Terms of Use — because we think you deserve to know this clearly before, not after, you rely on anything the Service tells you.

You are talking to AI

Every response you receive from an advisor profile on LifeChoices is generated by an artificial intelligence model. No human — not the mentor, not the public figure, not a LifeChoices staff member — is writing these responses to you in real time. We will never present an AI output as if it came directly from a human being.

What "interpretation" actually means

When you consult an advisor profile, the AI is drawing on:

• For personal mentors: the answers that mentor gave to our questionnaire, and any additional material they specifically authorized (see our Mentor Consent & Release);
• For public figures: publicly available material attributable to that person — their writing, recorded statements, interviews.

The AI uses that material to generate a response in the style and apparent values of the person modeled. It is doing pattern interpretation, not retrieval of an actual answer that person gave. If the mentor or public figure has never addressed your specific dilemma anywhere in the source material, the AI is extrapolating — making an educated guess consistent with the patterns it has — not reporting something that person actually said.

Specific limitations you should know about

• It can be wrong. AI models generate plausible-sounding text, not verified fact. A profile's response can misrepresent the real person's views, especially on topics the source material doesn't cover well.
• It can be outdated. A profile reflects the material available at the time it was built. People change their minds. Public figures' later statements may not be reflected in an older profile.
• It can sound more confident than it should. AI-generated text often reads as fluent and certain even when the underlying basis is thin. Confidence in tone is not evidence of accuracy.
• It is not a clinical, legal, or financial professional, no matter what profile you're consulting or what that real person's actual credentials are. See Section 3.2 of our Terms of Use.
• It does not know you the way a real mentor would. It only knows what you've told it in the current conversation and, where relevant, your stated dilemma. It has no independent knowledge of your life, history, or circumstances beyond what you provide.

What we do to reduce risk

• We label every advisor profile clearly as an AI interpretation, not a verified statement.
• We do not fabricate quotes and present them as things a real person definitively said.
• We design against "engineered warmth" — we do not tune the Service to maximize how much you like or trust it independent of whether it's actually helping you think clearly. See our Charter.
• Where a conversation suggests you may be in crisis, the Service may surface crisis resources. This is a safety measure, not a guarantee, and not a substitute for human help. See Section 9 of our Terms of Use.

What this means for you

Treat every response on LifeChoices as a perspective worth considering, generated by a tool — not as a verdict, and not as confirmation of what a real person would actually say or do. The decision, and the responsibility for it, stays with you.