AI Hub

Acceptable Use Policy (AUP)

Last updated: September 17, 2026

This Acceptable Use Policy ("AUP") defines the content standards, moderation mechanisms, reporting process and enforcement measures for all content generated, stored or distributed on AI Hub (the "Platform"), and applies to all users. By using the Platform you agree to comply with this AUP. It is part of our Terms of Service; where they conflict, this AUP prevails.

1. Scope

This AUP applies to all content forms created through the Platform:

  • Images: pictures and illustrations generated or edited by AI Sketch to Image and other tools;
  • Text: written content generated by AI Tarot Reading and other tools;
  • User input: prompts, tarot questions and other creative instructions you submit.

2. Content Standards and Prohibited Categories

The Platform prohibits using the service to generate, store or distribute the following six categories of content:

  • Sexual / NSFW content: any nudity, pornographic or sexually suggestive content, including such content rendered in an artistic style;
  • Violent / gory content: realistic or highly graphic violence, gore, abuse or self-harm, or content that glorifies violence;
  • Hate speech: content that attacks, demeans, discriminates against or incites hostility toward individuals or groups based on race, ethnicity, national origin, religion, gender, gender identity, sexual orientation, age or disability;
  • Child-unsafe content (CSAM): any sexually exploitative content involving minors, or any content that exploits or endangers minors;
  • Deepfakes / impersonation: generating or synthesizing identifiable real people without their consent, especially for deceptive, defamatory or sexual purposes;
  • Copyright / trademark infringement: content that infringes others' copyright, trademark, likeness or other intellectual property rights.

2.1 Zero tolerance (immediate enforcement)

The following are zero-tolerance violations. Upon discovery we remove the content and enforce against the account immediately, and reserve the right to report to the relevant authorities:

  • Child sexual abuse material (CSAM) or any sexual content involving minors;
  • Content promoting or facilitating terrorism, extremism or mass violence;
  • Instructions for creating weapons of mass destruction;
  • Content inciting genocide or ethnic hatred;
  • Non-consensual sexual deepfakes of real individuals.

2.2 Restricted content

The following is allowed only under the stated conditions; otherwise it is refused or removed:

  • Violent content: must have a clear creative or educational context and must not glorify real-world violence;
  • Politically sensitive content: potentially misleading content affecting public processes such as elections is restricted;
  • Medical / legal / financial advice: AI tarot readings and similar outputs are for entertainment only and must not be presented as professional advice.

2.3 General prohibited conduct

  • Generating or spreading misinformation or rumors;
  • Impersonating real individuals, brands or organizations;
  • Passing off AI-generated content as human-made to deceive others;
  • Bulk-generating spam or fraudulent content;
  • Using jailbreak prompts or other means to circumvent the Platform's safety mechanisms.

3. Content Moderation

The Platform uses a layered system of automated screening, human review and periodic sampling:

3.1 Automated screening (pre-generation and in-flight)

  • Input pre-screening: every generation request is first scanned by a content-safety scanning API (Waffo content safety check) before any billing or generation; prompts that fail the check never enter the generation pipeline;
  • Output filtering: image generation is constrained by the built-in content safety mechanisms of the official OpenAI API, and output that violates its usage policies is rejected upstream;
  • Usage-pattern analysis: request de-duplication, rate limits and quota controls monitor for abnormal calling behavior.

3.2 Human review

  • Requests blocked automatically but disputed, and content reported by users, enter a human review queue;
  • The image-mismatch dispute channel is handled by an administrator who compares each sketch with its result;
  • Reviewers must decide according to the category standards in this AUP, and handling records are retained.

3.3 Periodic sampling

We periodically sample generation records at random, evaluate and tune the accuracy of automated screening, and take retroactive enforcement against violating accounts.

3.4 Severity levels and response

  • L1 Critical (child safety, terrorism): blocked automatically; the account is enforced against immediately and reported as required;
  • L2 High risk (the six prohibited categories): blocked or removed automatically; the account is suspended pending review;
  • L3 Medium risk (general violations): removed after human review, with a warning;
  • L4 Low risk (minor violations): mostly report-triggered; warning and a request to modify.

4. Reporting

If you find content or behavior that violates this AUP, please report it to us:

  • Reportable situations: content violating the standards above, serious violations involving child safety, or AI impersonation infringing your own rights;
  • Reporting channel: submit through "Feedback & suggestions" on the creation page, marking it as a report and including the relevant access code, time and description;
  • Confidentiality: we keep reporter identities strictly confidential, and every report is reviewed independently and objectively;
  • Anti-abuse: malicious or false reports are recorded, and serious cases may result in loss of reporting privileges.

Response timelines

L1 critical violations receive an initial response within 6 hours and are handled within 24 hours; L2 within 24 hours and 3 business days; L3 within 3 business days and 7 business days; L4 within 5 business days and 15 business days. If more time is needed, we will proactively explain why.

5. Enforcement Measures

Depending on severity and level, we enforce along the following ladder:

  • Warning: a written warning requiring deletion or modification of the content (first minor violation);
  • Content removal: deletion of violating generated content and records (confirmed L3–L4);
  • Feature restriction: limiting available tiers or reducing call quotas (repeated minor or single L3);
  • Access-code suspension: temporary freezing of the access code and suspension of all services (L2 or accumulated violations);
  • Permanent ban: permanent freezing of the access code, with remaining credits forfeited and non-refundable (L1 or repeated L2);
  • Legal referral: suspected illegal activity is reported to law enforcement and investigated with their cooperation.

6. Appeals

If you disagree with an enforcement decision, you may appeal within 14 days of notice through "Feedback & suggestions" on the creation page. Appeals are reviewed by someone not involved in the original decision, usually with a conclusion within 10 business days. The original enforcement remains in effect during review.

7. Changes

We may update this AUP from time to time; updates are published on this page with the "Last updated" date, and material changes are highlighted on the site. Continued use of the service means you accept the updated policy. For questions about this AUP (including reports and appeals), contact us through "Feedback & suggestions" on the creation page.

Report or send feedback

Terms of Service