Dreamuna

AI Transparency

Last updated: 19 April 2026

This page describes how Dreamuna uses artificial intelligence to produce dream interpretations, articles, images, and audio narrations. It is the companion disclosure to our Privacy Policy and satisfies our labelling and information duties under Article 50 of the EU AI Act (effective 2 August 2026).

1. What Dreamuna is, in AI terms

Dreamuna is an AI-assisted dream reflection service. Everything you receive from us — the written interpretation of your dream, the chat replies from the Dreamuna persona, the long-form symbol articles, the painted dream-scene images, and the spoken narration — is generated by large language and multimodal models, with light structural templating on top. There is no human therapist or clinician reading your dream. Our role is to design the prompts, the safety layer, and the user experience around these models.

2. Which models we use

Generated output is produced by the following third-party AI models:

  • Dream interpretation (partial + full) — OpenAI GPT-4o
  • Dreamuna chat replies — OpenAI GPT-4o mini
  • Long-form symbol articles on /dream-about/ — OpenAI GPT-4o
  • Dream-scene images — OpenAI gpt-image-1
  • Audio narration of readings — OpenAI gpt-4o-mini-tts

All of the above are served by OpenAI as a sub-processor. Dreamuna does not train any of these models on your content. Sub-processor details and data-transfer safeguards are listed in our Privacy Policy.

3. What AI can and can't do for you

What Dreamuna's AI is genuinely good at:

  • Naming patterns — picking up recurring symbols, emotional tones, and dream structures you may not have articulated.
  • Offering frames — psychological lenses (Jungian, emotion-regulation, narrative) to read your dream against.
  • Prompting reflection — surfacing journaling questions and waking-life cues tied to what you described.

What AI cannot do — and what we explicitly don't claim:

  • It is not therapy. It is not a medical device, not a diagnosis, and not a substitute for a qualified clinician.
  • It can be wrong. Large language models produce plausible-sounding text that may be inaccurate, speculative, or inconsistent between readings.
  • It has no memory of you. Each reading is produced from the dream you submit plus the minimum context required (locale, intensity sliders, chat history for the same dream). It has no access to your other dreams unless a specific feature says so.
  • It cannot predict the future. Dream interpretation here is a reflective exercise, not divination.

4. How we label AI-generated content

Wherever AI produces user-visible content, we label it as such. You will see:

  • A badge below every full and partial dream interpretation.
  • A caption under every generated dream-scene image ("AI-generated image").
  • A note at the top of every long-form symbol article on /dream-about/.
  • A disclosure line in every interpretation email we send and in every downloadable PDF.
  • A notice next to the Dreamuna chat so replies are clearly identified as AI output.

5. Your rights under Art. 50 and related EU law

You have the right to know when you are interacting with an AI system; this page and the badges across the product satisfy that right. You also retain every right granted to you under the GDPR — access, rectification, erasure, restriction, portability, objection — as described in our Privacy Policy. If you believe AI-generated content about you is inaccurate, misleading, or inappropriate, email us at hello@dreamuna.com and we will act on it.

  • Right to know — every AI-generated surface is labelled.
  • Right to opt out — you can stop using Dreamuna at any time, cancel your subscription, and request deletion of your dreams and chat history.
  • Right to correction — if a reading contains factual errors about you, contact us and we will delete or amend it.
  • Right to human review — you can email us with any AI-generated content you want a human to look at; we will reply within five business days.

6. Contact

Questions about AI use, model choice, labelling, or your rights: hello@dreamuna.com. We reply in the language you write to us in where we can.

See also: Privacy Policy · Terms of Service · Contact