Tarosyn Legal

Responsible AI Principles

Last updated May 31, 2026

LEGALLY REVIEWED

Tarosyn uses artificial intelligence to power tarot readings, astrological insights, Mystic Personas, dream interpretation, and more. We believe that AI can be a genuine tool for self-reflection and personal growth — but only if it is built and deployed responsibly. The four principles below describe the standards we hold ourselves to as we develop, deploy, and iterate on every AI-powered feature inside the app.

PRINCIPLE 1

Fairness

We design our AI systems to serve all seekers equitably, regardless of background, identity, or belief.

Card interpretations, horoscopes, and Oracle responses are crafted to be culturally aware and free from stereotyping based on gender, race, religion, or any protected characteristic. Our Mystic Personas are fictional archetypes — not representations of any real-world culture or faith tradition — and are clearly presented as such. We actively monitor AI output for patterns that could disadvantage or misrepresent particular groups, and we revise prompts and fine-tuning data when issues are found. Subscription tier does not affect the quality or impartiality of AI reasoning; premium features add breadth, not preferential treatment.

PRINCIPLE 2

Safety

We build safeguards that prevent our AI from causing harm — and we tell users clearly when they should seek human help instead.

Every AI surface carries a disclosure that readings and interpretations are for personal reflection only and do not constitute medical, psychological, legal, or financial advice. Content filters and system-level guardrails block the generation of harmful, violent, sexually explicit, or crisis-triggering content. When a user's input suggests they may be in distress or crisis, our AI is instructed to respond with empathy and direct them to appropriate professional or emergency resources — not to continue the reading as normal. Rate limits, output size caps, and per-feature usage controls are in place to prevent AI resources from being abused in ways that could harm the service or other users. We conduct internal safety reviews before releasing new AI features or significantly changing existing prompts.

PRINCIPLE 3

Transparency

We tell users what AI is doing in plain language — without obscuring the technology behind mystical framing.

Every reading, persona message, and AI-generated image is clearly attributed to AI generation so users understand the source of the content. Our AI Transparency page explains which foundation models we use, what data is sent to them, and what limits we place on that use. The Truth State dial and AI verbosity controls give users direct, understandable influence over how AI output is styled — putting them in the driver's seat. We do not use AI to create the impression of human moderation, human authorship, or human relationships without explicit disclosure. When AI features are experimental or in beta, we say so prominently within the feature itself.

PRINCIPLE 4

Accountability

We own our AI's outputs and maintain clear processes for identifying, addressing, and learning from failures.

A named internal team is responsible for reviewing AI-related policy, safety incidents, and principle adherence on an ongoing basis. Users can report AI-generated content they believe to be harmful, inaccurate, or biased through the same in-app Report flow used for user-generated content. We log AI safety incidents — including any cases where a user's input triggered a crisis-response path — and use them to improve our systems. We will not grant third parties access to user reading data or conversation history for AI training purposes without explicit, informed consent. Material changes to how AI is used in the app will be communicated to users in advance, not buried in a terms update. Scope & limitations

These principles apply to all AI-powered features within the Tarosyn application — including readings, the Oracle, Mystic Personas, Dream Interpreter, Photo Reading, AI card generation, and AI-assisted blog content. They do not govern third-party model providers (OpenAI, etc.) beyond the contractual and technical controls we place on our API usage.

We acknowledge that AI systems can produce unexpected or imperfect results. No principle statement eliminates that risk entirely. Our commitment is to continuous improvement, honest disclosure when things go wrong, and a genuine effort to do better.

Questions & feedback

If you have a concern about AI behaviour in Tarosyn — a response that felt biased, harmful, or inaccurate — please reach out at [email protected]. We read every message and use them to improve the product.

AI Transparency AI Safety Policy AI Bias Report Acceptable Use Policy

Questions? [email protected]