Legal

AI Transparency Notice

Last updated: 7 September 2026

This notice explains how we use artificial intelligence on the Platform, what it does, what it cannot do and what this means for you. We provide it having regard to Regulation (EU) 2024/1689 (the Artificial Intelligence Act).

In short

  • We use artificial intelligence in three places: the written interpretation in your Architecture Report, your weekly Coaching insight, and the AI Companion (the chat feature).
  • Everything else you see is calculated, not written by AI. Your axis scores, your structural classification, your drift results and your compatibility results are worked out from your own answers to the Assessment (the questionnaire you complete when you join) using our own framework. The same answers always give the same result.
  • When you use the AI Companion, you are talking to an AI system, not to a person.
  • Nobody at TAW reads or checks the AI-written content before you see it. It can be wrong, incomplete or inconsistent, and it may differ from one session to the next.
  • Nothing on the Platform, whether AI-written or calculated, is medical, psychological or psychotherapeutic advice, and none of it is a diagnosis.

The sections below set all of this out in full.

1. Who We Are

THE ARCHITECTURE WITHIN S.R.L., a Romanian company with its registered office at Bucharest, 122 Mihai Bravu Road, Building D27, 1st Entrance, 2nd Floor, Apartment 6, 2nd District, Romania, registered with the Trade Registry under no. J2026052769003, having tax identification number 55529344 (“TAW”, “we” or “us”), operates The Architecture Within platform at thearchitecturewithin.com(the “Platform”).

If you have a question about anything in this notice, you can contact us at admin@thearchitecturewithin.com.

2. Where We Use Artificial Intelligence

We use artificial intelligence to produce three things:

  • the written interpretation in your Architecture Report — the identity, stress signature, cognition, emotional, relational and growth sections. This is generated once, when you complete your Assessment, and then saved;
  • your weekly Coaching insight, a short reflective summary of the journal entries you wrote that week; and
  • the replies you get from the AI Companion.

We do not use artificial intelligence for anything else. In particular, these are calculated from your answers, not written by AI:

  • your four axis scores;
  • your structural identity and pattern code;
  • your drift classifications;
  • your compatibility scores and your Relationship Intelligence report; and
  • your first-moves guidance and your Wall cards.

Those results come from arithmetic applied to your own answers, and from text we wrote in advance and select using fixed rules. Nothing about that process learns or changes on its own. If you answered the Assessment the same way twice, you would get the same result both times.

Assessment, Architecture, Coaching and AI Companion are the names we use for these features on the Platform and in our Terms of Use. They carry the same meaning here.

3. The AI Companion

The AI Companion is a conversation with an AI system. You are not talking to a person, and no person is reading the conversation as it happens.

It uses what you write in the conversation together with information from your Assessment and Architecture, so its answers refer to you specifically. That does not make them more reliable. An answer that sounds personal is still generated automatically, and it has not been checked by anyone.

If you write something suggesting you are in crisis, the AI Companion does not answer it. The message is not sent to the AI at all. You get a fixed reply from us instead, pointing you to help. That reply is not an assessment of your situation, and we do not monitor these conversations.

The AI Companion is not a professional, and its answers are not a professional opinion.

4. What AI-Generated Content Is, and What It Is Not

Where content on the Platform is written by artificial intelligence, we label it. The labels are:

  • in the AI Companion, a notice that stays at the top of the page: “You are chatting with an AI system. Responses are AI-generated from your profile data and are not professional, medical, or psychological advice.”;
  • in your Architecture Report, where the written interpretation was generated by AI, a notice reading: “The written interpretation in this report is AI-generated from your measured results. It is a reflective tool, not professional, medical, or psychological advice.”; and
  • on your weekly Coaching insight, the label “AI-generated insight”.

Your Report does not always use AI-written text. If the generated version does not pass our checks, the Report shows text we wrote in advance instead, and the notice above is not displayed.

Your scores, classifications and compatibility results carry no AI label, because there is no AI in them.

Where content is written by artificial intelligence, it has real limits. What it produces may be:

  • wrong;
  • incomplete;
  • different from one session to the next, even on the same question;
  • based on assumptions about you that are not correct;
  • unsuitable for your situation; or
  • affected by bias in the underlying technology or in the information given to it.

The fact that AI-written content is personal to you, or refers to your Assessment or your Architecture, does not mean it has been reviewed or verified by anyone. It is generated text, not a verified conclusion about you.

You decide what to do with it. Please think about whether what you are reading actually fits your situation before you rely on it or act on it.

5. This Is Not Medical or Psychological Advice

The Platform uses artificial intelligence within our own framework for describing how people think, decide, relate to others and respond under pressure. That framework is not a clinical one.

Nothing on the Platform is:

  • a medical or psychological diagnosis;
  • a psychological or psychometric assessment;
  • psychotherapy or counselling;
  • medical or psychiatric treatment;
  • professional healthcare advice; or
  • any other regulated professional service.

If you need advice or help with a medical, psychological or other professional matter, please speak to a qualified professional.

6. If You Are Struggling

If you are going through something difficult, the AI Companion is not the right place to take it. It cannot assess how you are, it cannot tell how serious something is, and it is not a substitute for help from a person.

If you are in distress, or you are worried about your safety or someone else’s, please contact an emergency service or a qualified professional. In Romania the emergency number is 112.

If you write something to the AI Companion suggesting you are in crisis, we show you emergency and helpline information at that point. We keep that information current, but we do not run those services, we cannot reach them on your behalf, and we do not monitor whether you contact them.

7. Your Personal Data

Our AI features process personal data that you give us or that is generated as you use the Platform. Depending on the feature, this includes your Assessment answers and results, information relating to your Architecture and, for the AI Companion, what you write in the conversation.

Please keep in mind that whatever you write to the AI Companion is processed by us and by the provider of the underlying technology. Try not to share anything you would not want processed in that way.

Our Privacy Notice explains what personal data we collect, why, on what legal basis, who receives it, whether it leaves the EU, how long we keep it and what rights you have.

8. Human Review and Decisions About You

We do not review AI-generated content before you see it.

We do not use artificial intelligence to decide whether you can use the Platform, to suspend or close your account, to penalise you, or to take any other decision that has legal effects for you or that affects you in a similarly significant way.

AI-generated content supports the features of the Platform. It does not replace your own judgment, and it does not decide anything about you.

9. The Technology Behind It

The three AI features described in section 2 are built on technology supplied by a third party, currently Claude, provided by Anthropic, PBC. The calculated parts of the Platform do not use it.

We set our own instructions, restrictions and limits on how that technology is used on the Platform. Its behaviour also depends on characteristics and limitations of the underlying technology, which are set by the provider and are not within our control.

Using third-party technology does not change anything in this notice. The features you use on the Platform are ours, and so is our responsibility for them.

10. Changes to This Notice

We will update this notice when we change the Platform, the AI features, the technology behind them or the way any of it works, and where the law requires it.

The current version is always available on the Platform and shows the date it was last updated.