How your coach works, what it can do, what it cannot do, and what biases to watch for.
It is not a human. It does not have feelings, memory between sessions beyond what you give it, or lived experience. It is a powerful, careful, language-pattern system that, with the right scaffolding, can hold a thoughtful coaching conversation. We wrote the instructions it follows. The wisdom in the room is yours.
Inside Being Me, anywhere you see the name My Coach, its speech-bubble mark , or the small badge that reads AI coach, you are interacting with an AI.
Specifically, you are interacting with a large language model: an AI trained on a very large body of text to read and write language. We at Being Me have written a detailed set of coaching instructions (what's called a "system prompt") that shapes how it responds to you. Those instructions include our coaching philosophy, the frameworks Being Me uses (REACH, PACTS, SMARTER), the response structure we want, and the safety rules the AI must follow.
So while the language model underneath is general-purpose, the coach you experience is purpose-built for Being Me.
What you send the coach is used to write its reply. It is not used to train the AI provider's models. The Privacy Policy names that provider and says where your data goes.
This is a simplified explanation. It is not technically precise. It is meant to give you an honest feel for what the AI is doing.
A large language model is, at its core, a very sophisticated pattern matcher. It has been trained on an enormous amount of text (books, articles, websites, code, conversations), and from all that text it has learned the statistical patterns of how human language works. When you send it a message, it doesn't "think" the way you do. It generates a response one word at a time, each word chosen because it is the most likely continuation given everything that came before.
What makes modern LLMs feel intelligent is that "the most likely continuation", across enough patterns, enough context and a careful enough system prompt, turns out to capture an enormous amount of what we call reasoning. It is not the same as human thought. But it is not nothing, either. The honest description is somewhere in between, and the boundary is genuinely fuzzy.
Two implications for how to use it well: first, the AI is only as good as the context you give it: the more honest and specific your check-ins, values, and Why, the more relevant the coaching will be. Second, the AI can be confidently wrong. Pattern-matching does not equal truth-finding. When the stakes are high, verify.
If you are in a moment where you need any of the things in the right-hand column, please reach out to someone qualified. The AI coach is a thinking partner, not a replacement for the humans who matter in your life.
Every AI model trained on human-generated text carries the biases of that text. Pretending otherwise would be dishonest. Here are the specific biases we know about and that you should keep in mind when reading any coaching response.
This list is not exhaustive. The most reliable single safeguard against AI bias is your own self-knowledge, which, not coincidentally, is what Being Me is designed to build.
The AI coach reasons about your situation using four interlocking frameworks. We developed these for Being Me. Here is what each one is, with a candid note on its evidence base.
The AI coach operates under explicit safety rules. The most important is what we call Safety Rule 0, written into the AI's system prompt:
This rule sits on top of every coaching exchange. Alongside it, the Being Me app itself scans your messages for language that indicates distress, before the message ever reaches the AI. When that scan triggers, the app shows you a moment-of-care message with crisis resources, rather than proceeding with coaching.
We will be the first to say that no automated safety check is perfect. We are continuing to improve this layer. It is the highest-stakes part of the platform and we treat it that way. If you ever feel the safety check missed something it should have caught, please tell us at support@beingme.me.
When the scan triggers, the app keeps a short record on your device that it happened: how serious, and what kind. Never your words. It stays in your browser and is not sent to us. It is deleted after 12 months, and when you sign out. The privacy policy has the detail.
To repeat the most important thing on this page, because it matters more than the rest:
Being Me is not a therapist, counsellor, psychologist, psychiatrist, doctor, lawyer, financial advisor, or crisis line. It is a personal development coaching platform.
If you are in distress, in danger, in crisis, or facing a serious medical, legal, or financial decision, please reach out to a qualified human professional. We have listed the most-used crisis lines below. They are also available inside the app under "A moment of care".
If you would like to work with a credentialled human coach, you can find one through the International Coaching Federation directory.