AI Acceptable Use Policy

1. Scope and current availability

This policy is part of the Terms of Service of PhotoCraft Online, operated by Tengzhou Lingshun Yuanshi Jiaoyu Zixun Gongzuoshi. It applies to prompts, submitted reference images, Agent requests and generated images handled by our optional online AI services. It does not imply that we scan documents edited only on your device.

Paid AI generation is not currently enabled. The screening process below is a condition of activation, not a claim that an automated moderation service is already deployed. We do not offer public image galleries, audio generation or video generation.

2. Prohibited content — six explicit categories

Pornography / NSFW: do not generate explicit nudity, sexual acts, sexual exploitation or non-consensual intimate imagery.

Violence / gore: do not generate graphic injury, dismemberment, torture, sadistic violence, or content promoting terrorism, extremism or mass violence. Instructions for weapons of mass destruction are prohibited.

Hate speech: do not generate dehumanization, slurs, threats, incitement or discriminatory abuse targeting people because of race, ethnicity, religion, nationality, sex, gender identity, sexual orientation, disability or another protected characteristic.

Child safety / CSAM: any sexualized depiction of minors, child sexual abuse material, grooming, exploitation or instructions facilitating harm to children are strictly prohibited, including synthetic content.

Deepfakes / impersonation: do not fabricate a real person's identity, face, intimate images, statements or actions, whether or not that person consents. Do not impersonate people, brands or organizations, commit fraud or present generated media as authentic evidence.

Copyright / trademark infringement: do not submit or generate content that infringes copyright, trademarks, privacy, publicity or other rights. Only upload your own images or material you are authorized to use; permission to upload does not permit prohibited deepfakes.

3. Upload permissions and generated content

Upload only images you own or are authorized to use, including permission from people depicted. Permission does not override the prohibition on real-person deepfakes. Do not upload confidential material you are not authorized to share.

Do not pass generated content off as authentic evidence or impersonate another person or organization. AI results are identified as generated in the AI workflow. Review outputs before use; you remain responsible for lawful publication and third-party rights.

4. Input and output moderation

Before paid AI is activated, submitted prompts and reference images must be screened before model submission, and generated text and image outputs must be screened before delivery. Unsafe inputs are rejected; unsafe outputs must be withheld. Provider restrictions must remain enabled and may not be bypassed. An unavailable or inconclusive safety check must block or hold the request rather than deliver unchecked content.

The independent automated screening provider has not yet been activated. Provider safety filters are an additional layer and do not replace our required input and output checks. AI remains unavailable for paid use until that screening, blocking and audit workflow is implemented and tested.

The operator manually reviews reported content and available task records, prioritizes child safety and credible threats, and assesses context before deciding on removal or restrictions. We do not claim a 24-hour staffed moderation team. Review access is restricted to information needed for the investigation.

At launch, the operator must test blocking behavior, review incidents and revise safety rules when failures are identified. Pricing or purchase availability must not bypass these launch conditions.

5. Report unsafe content or infringement

Email contact@photocraft.digital with the subject 'Content report'. Include a task or order identifier or relevant URL, the category of concern and a short explanation. Rights holders should identify their work or mark, the allegedly infringing result and their contact details. Do not attach CSAM, illegal intimate imagery, card details or unnecessary personal data.

For credible immediate danger, contact the appropriate emergency authority; email support is not an emergency service. We prioritize child safety, exploitation and credible-threat reports. We aim to acknowledge other reports within 2 business days; investigation times depend on available evidence. Reporter information is restricted to those handling the case, subject to lawful disclosure obligations.

6. Review, enforcement and appeals

Reports are received by the operator, triaged for urgency, reviewed against these categories and relevant records, and followed by a documented decision. We may request limited additional information and notify the affected user or reporter where appropriate and lawful.

Actions may include a warning, rejected generation, withholding or deleting a hosted result, restricting AI or wallet access, suspending abusive sessions or linked identities, and terminating service for serious or repeated breaches. Where legally required, relevant records may be preserved and reported to competent authorities. We cannot remove copies already downloaded or control unrelated third-party websites.

Restrictions do not automatically forfeit prepaid balances. Refunds and billing corrections follow the Terms of Service and mandatory consumer rights. Confirmed model usage and unused reservations are handled under the published credit rules.

To appeal, email contact@photocraft.digital with the subject 'Content appeal', the task identifier and why you believe the decision was incorrect. The operator will reassess the decision and explain the outcome where permitted. Do not evade a restriction while an appeal is pending.