AI Disclaimer

1. What this covers

This disclaimer applies to every AI feature on hardiktrehan.com and the HT Hub sub-site, including but not limited to: the AI Battle Arena (debate, tournament, voice narration), the consolidated astrology and spiritual suite (Kundli, Astrologer, Predictor, Dream Interpreter, Tarot, Numerology, Compatibility, Past Life, Lucky Number, Career, Vibe Match), Prompt Lab, Image Prompt, Code Duel, Roast Battle, Model Showdown, and any future AI-powered tool we add.

These tools route your inputs through third-party AI models via OpenRouter (Google, OpenAI, Anthropic, Meta, Mistral, DeepSeek, xAI, Moonshot, Qwen), generate text-to-speech via ElevenLabs for paid users, and pull web research via Perplexity in the Arena. See the Data Processing Addendum for the full subprocessor list.

2. Entertainment, not advice

The AI tools exist to be fun, novel, and occasionally insightful. They are not a substitute for:

• A licensed physician, therapist, psychologist, or psychiatrist.
• A licensed attorney or legal aid organization.
• A licensed financial advisor, accountant, or tax professional.
• A certified astrologer, tarot reader, numerologist, or other practitioner.
• A career counselor, life coach, or relationship therapist.
• An emergency service, crisis hotline, or safety professional.

Treat AI output as you would a friend who reads a lot but has no credentials — entertaining, sometimes thought-provoking, sometimes wrong about the details. Verify anything important.

3. AI hallucinations are real

Large language models generate text by predicting the most likely next token, not by retrieving verified facts. As a result, they can — and do — produce content that sounds authoritative but is factually wrong, including:

• Fake citations, court cases, studies, or news events.
• Wrong dates, numbers, addresses, statistics, and quotes.
• Invented “facts” about real people, places, and products.
• Plausible but incorrect medical, legal, or scientific claims.

This is true of every commercial AI model on the market today, including the most expensive ones. Verify any factual claim from AI output against a primary source before relying on it.

4. Astrology & spiritual tools specifically

We compose readings by combining classical astrological / symbolic frameworks with AI text generation. The output is the AI’s interpretation of those frameworks; it is not produced by a human astrologer, and the AI may misapply the rules.

Specifically, you should not:

• Make medical decisions (start, stop, or change treatment) based on a reading.
• Make legal or financial decisions (sign contracts, invest, lend) based on a reading.
• End or start a relationship, marriage, or business partnership solely on a compatibility score.
• Hire, fire, promote, or demote anyone based on a reading.
• Bet, gamble, or trade based on “lucky numbers” — they are random text.
• Believe a “past-life” narrative as historical fact about you or anyone else.

If a reading tells you something alarming about your health, a person you love, or your future — pause, breathe, and talk to a qualified human. The reading is a story, not a diagnosis.

5. Mental-health and crisis content

The AI tools are not mental-health professionals and cannot provide therapy, diagnosis, or crisis intervention. If you are experiencing thoughts of self-harm, suicide, or an acute mental-health crisis, please contact a crisis service:

• United States: 988 Suicide and Crisis Lifeline (call or text 988).
• United Kingdom: Samaritans, 116 123.
• Canada: Talk Suicide Canada, 1-833-456-4566.
• Australia: Lifeline, 13 11 14.
• India: iCall, +91 9152987821; Vandrevala Foundation, 1860-2662-345.
• Other regions: findahelpline.com lists vetted crisis lines worldwide.
• Immediate emergency: call your local emergency number (911, 112, 999, etc.).

6. Medical, legal, financial, professional advice

Nothing on the Site — including any AI-generated text — is medical, legal, financial, tax, psychological, psychiatric, accounting, investment, or any other kind of professional advice. No professional relationship is created by your use of the Site.

• Medical questions: see a physician or licensed clinician.
• Mental-health questions: see a licensed therapist or psychiatrist.
• Legal questions: see an attorney licensed in your jurisdiction.
• Financial / investment questions: see a credentialed financial advisor.
• Tax questions: see a CPA or tax attorney.

7. Code Duel — code is a starting point

Code generated by Code Duel (or any other AI tool on the Site) is not production-ready. Review every line. Run it in a sandbox before deploying. Test it. Audit it for security vulnerabilities (SQL injection, XSS, insecure deserialization, hardcoded secrets, weak crypto, missing input validation). The AI may invent functions, APIs, library names, and version numbers that do not exist. The AI may produce code that compiles but silently corrupts data or leaks credentials.

You are responsible for the code you ship.

8. AI Battle Arena debates — synthetic, not authoritative

AI Battle Arena debates are synthetic conversations between AI models. The participants are not experts, journalists, scientists, or eyewitnesses — they are language models generating plausible-sounding arguments. Their citations may be fabricated; their statistics may be invented; their rhetorical confidence is not evidence of correctness.

Treat the debates as a thinking exercise or entertainment, not as a source of truth on contested topics. For real journalism, see our news partner vyraa.com or any other reputable outlet that practices verification.

9. Bias, offense & cultural assumptions

AI models are trained on data scraped from the internet and inherit biases from that data. Outputs may reflect or amplify stereotypes about race, ethnicity, religion, gender, sexual orientation, body type, language, disability, age, or nationality. Outputs may be culturally Western-centric or English-centric even when you ask in another language.

If you receive an output that is biased, offensive, or contains hate speech, please report it to abuse@hardiktrehan.com with the prompt and the output. We use reports to refine prompts, add guardrails, or change models.

10. Privacy — treat inputs as public

When you submit a prompt, debate topic, code snippet, dream, or any other AI input, that content is transmitted to one or more third-party AI providers via OpenRouter (and to ElevenLabs for voice and Perplexity for web lookups). We rely on those providers to honor their privacy commitments, but we cannot guarantee what they do with it. Do not enter sensitive, confidential, regulated, or personally identifying information into the AI tools beyond what is strictly required for the reading (e.g. a birth date for astrology).

See the Privacy Policy for the full data flow.

11. Minors

The Site is not directed to children under 13. The AI tools can produce content unsuitable for minors. Parents and guardians: see our Children & COPPA Policy.

12. Limitation of liability

By using any AI tool on the Site, you agree that the Site is provided “AS IS,” that AI outputs are not endorsed by us, and that our total aggregate liability for any claim arising from your use of any AI tool — including any decision you made based on AI output — is limited as described in section 11 of the Terms of Service. Nothing here limits liability that cannot be limited under applicable law.

13. Emergency & safety resources

If you are in immediate danger, call your local emergency number first. For crisis-line numbers by country, see section 5 above or findahelpline.com.

14. Contact

Questions about this disclaimer: legal@hardiktrehan.com. Reports of harmful or biased output: abuse@hardiktrehan.com.