AI Content Policy

This policy applies to every image generated with Jewelry AI Studio ("the Service"), operated by HAOWX. It forms part of our Terms of Service and is linked from the footer of every page. The Service is a product-photography tool that replaces the background of a jewelry photo you upload; it is not a general-purpose image generator and must not be used as one.

1. Prohibited content categories

You must not upload, request, generate, publish or distribute anything falling into the following six categories. These rules apply to the photo you upload, the parameters and text you supply, and the generated result.

1. Pornography / NSFW — nude, sexually explicit or sexually suggestive content, sexualised body parts, fetish content, escort or adult-service material, and links to such content.
2. Violence / gore — graphic injuries, blood, mutilation, torture, corpses, weapons used to threaten, or content that glorifies violence, self-harm or cruelty to people or animals.
3. Hate speech and harassment — slurs, extremist or hate symbols, and imagery that attacks, demeans, threatens or incites discrimination or violence against a person or group because of race, ethnicity, nationality, religion, gender, sexual orientation, disability, age or similar traits.
4. Child sexual abuse material (CSAM) / unsafe content involving minors — any sexualised depiction of a person who is or appears to be under 18, grooming or exploitation material, and any use of a minor's likeness. The Service is for adults only; imagery of minors must not be generated at all.
5. Deepfakes / impersonation — face swaps, synthetic likenesses of real people, fake identities or forged documents, fabricated endorsements or testimonials, impersonation of a brand or public figure, or presenting an AI-generated person, product or certificate as an authentic photograph or third-party endorsement.
6. Copyright / trademark infringement — counterfeit branded goods, another brand's logo, protected designs or characters passed off as your own, and any third-party work used without the rights holder's permission.

Also prohibited: illegal goods or services (including weapons, drugs and wildlife products), defamation, fraud, attempts to bypass our content moderation, scraping the Service, reselling raw access to the underlying model, or using it to build a competing dataset.

2. Platform content moderation mechanism

We operate a two-stage review for each generation. Both stages are logged and are designed to fail closed: if a check cannot run, the request is held for review rather than delivered unmoderated.

• Before generation. The prompt and request parameters are screened against a blocked-term list and an automated prompt-safety service (Waffo Prompt Sift). Only requests marked "allow" may be charged and sent to the image model. Requests marked "block" or "review" are refused, no credits are deducted, and no model call is made.
• After generation, before display. The result must pass an output check (result-format validation and an image-moderation step, or a second prompt-safety scan where an image endpoint is not configured). A result that does not pass is not shown to the user, is marked failed, and its credits are returned automatically.
• Audit log. Every allow / review / block verdict is recorded with the generation ID, account, stage, method, reason code, matched category and a hash of the prompt (never the raw prompt). Logs are retained in line with our Privacy Policy and are used to investigate reports and repeated abuse.
• Human review. Automated verdicts are not the final word: reports and flagged content are reviewed by a human operator before enforcement.

3. User reporting channel

Anyone may report content that appears to breach section 1 — an account is not required. Use the "Report content" link in the footer of every page, or email wx975lm@163.com with the subject line "AIGC Content Report". Reports are accepted in English and Chinese.

To help us act quickly, please include:

• the image URL or a screenshot, and when you saw it;
• the generation ID from your image history, if available;
• which prohibited category it concerns and a short description;
• an email address if you want a reply (anonymous reports are accepted, but we cannot tell you the outcome).

Never attach CSAM or other illegal imagery to a report. Send only a URL or a text description; we will collect and preserve evidence lawfully. Reporter identities are kept confidential.

4. How reports are handled

1. Acknowledgement — we confirm receipt within 1 business day.
2. Verification — we check the generation record, the moderation log for both stages and, while the temporary link is valid (about 24 hours), the result itself. Where needed we ask the reporter or the rights holder for more detail.
3. Decision and action — normally within 3 business days, and no later than 7 business days where a third party must be consulted. Depending on the finding we may disable or remove the download link and stored result, mark the generation failed and return its credits, remove the content from any public surface, warn, restrict, suspend or terminate the account, and preserve evidence.
4. Urgent cases — reports involving CSAM, credible threats or imminent harm are actioned within 24 hours (immediate link and account blocking) and reported to the competent authority where the law requires it.
5. Appeals — a user whose content or account was restricted may appeal to the same address within 7 days; we conduct one re-review and reply within 5 business days.
6. Records — the report, findings, action taken and timestamps are logged alongside the moderation records described in section 2.

5. AI-generated content and your responsibilities

Results are AI-edited images, not untouched photographs. Do not present them as unaltered photographs, appraisal certificates or third-party endorsements, and disclose AI editing where a marketplace, platform or law requires it. You are responsible for the photos you upload, for having the rights to them, and for how you use the images you generate.

6. Contact

Support, refunds, data requests and content reports: wx975lm@163.com. We answer general enquiries within 3 business days; the shorter deadlines in section 4 apply to content reports.