Acceptable Use Policy

Last updated: 2026-06-02 · Version 1.1

This AUP supplements the Terms of Service. Violations may result in immediate account suspension and forfeiture of remaining credit.

Strictly prohibited (zero tolerance)

  • Generation, distribution, or facilitation of child sexual abuse material (CSAM). Detected attempts will be reported to authorities.
  • Content depicting real, identifiable individuals in sexual, violent, or defamatory contexts without their consent (non-consensual deepfakes).
  • Content intended to incite violence or hatred against persons or groups based on race, religion, ethnicity, gender, sexual orientation, or other protected characteristics.
  • Content facilitating terrorism, mass casualty events, or weapons of mass destruction.
  • Fraud and impersonation: phishing pages, fake government documents, forged identity documents, fake bank/payment messages.
  • Coordinated disinformation intended to influence elections or financial markets.

Restricted use

The following are not banned outright but require special care. Repeated violations may result in suspension:

  • Medical advice / diagnosis, output must not be presented as professional medical opinion. Always include a disclaimer pointing users to licensed clinicians.
  • Legal advice, same disclaimer requirement; refer users to advokat/notaris.
  • Financial advice, same; refer to OJK-licensed advisors.
  • Adult content (involving consenting adults only), allowed for personal use but must be age-gated if redistributed.
  • Political content, allowed for educational and journalistic purposes; must not impersonate candidates.

Technical abuse

  • No denial-of-service attacks, scraping at rate exceeding your tier limits, or attempts to enumerate other users' keys.
  • No attempts to extract training data, model weights, or proprietary infrastructure information.
  • No automated multi-account signups to claim free credit ("credit farming"). One real user, one account.
  • No proxying/reselling the API as your own product without entering a reseller agreement.
  • No circumventing model safety features through adversarial prompting in ways that produce content matching the prohibited categories above.

Reporting violations

If you encounter content or behavior violating this AUP, report to hello@epithre.com with subject "AUP report". Include evidence (request IDs, timestamps) where possible.

Enforcement

Epithre reserves the right to suspend accounts immediately and without prior notice for serious violations. Less severe violations will receive a warning first. We log abuse signals (rate-limit hits, high-error patterns) and review them regularly.