Acceptable Use Policy

Scope and seven prohibited categories

This policy covers prompts, negative prompts, uploaded images/audio, generated images/videos and content stored or shared through Dola AI. Do not request, upload, generate or distribute:

1. Pornography / NSFW: sexually explicit material, sexual exploitation, or nudity intended for sexual gratification.
2. Violence / gore: graphic injury, mutilation, extreme brutality, or promotion/facilitation of violent harm.
3. Hate: attacks on protected groups, dehumanization, or incitement to discrimination or violence.
4. Unsafe child content: child sexual abuse material (CSAM), sexualized minors, grooming, exploitation or other content endangering children. This is a zero-tolerance category.
5. Deepfakes / impersonation: unauthorized synthesis of a real person's likeness/voice, deceptive identity substitution, fabricated endorsements/evidence, or falsely presenting synthetic conduct as authentic.
6. Non-consensual intimate imagery (NCII): real or synthetic intimate content without the subject's explicit consent, including threats to create or publish it.
7. Copyright / trademark infringement: unauthorized protected content, misuse of logos/brands or other infringement of third-party intellectual-property rights.

Fraud, harassment, credible threats, terrorism, illegal activity, malware, safety bypasses and unauthorized access are also prohibited. Do not claim Dola AI is an official model-provider application. Disclose synthetic content where appropriate; never present fabricated evidence as real.

Real-person authorization

Upload only material you own or have authority to process. Each identifiable person must have consented to the intended AI use of their photo, likeness or voice. Public availability or possession of a file is not permission. Unauthorized photos/voices and non-consensual intimate material are prohibited. The upload/generation declaration, where requested, is an attestation, not independent verification. We may request authorization evidence and restrict disputed content. Do not upload credentials or confidential documents.

Content safety mechanisms

For Waffo-enabled generation, the final text prompt, fixed effect prompt and any negative instructions are screened on the server using Waffo Prompt Sift before a job or credit charge. Only a valid allow verdict permits generation. A block or review verdict prevents the request from starting. If screening is unavailable, incorrectly configured or returns an invalid response, the request is temporarily refused. These refused requests do not create a generation job or charge generation credits. A review verdict does not automatically queue a generation to start later; revise the prompt or contact support. Model switching and direct API requests do not bypass the same server check.

Manual review of stored or reported outputs is separate from pre-generation text screening. An existing output flagged for operator review may remain visible and downloadable to its owner until a removal decision is made. The operator inspects the complete image or video and records an approval or rejection reason. Rejected results are removed, remain inaccessible, and deducted generation credits are returned. This is manual post-publication review, not an automated image/video-scanning claim.

Models may apply additional safeguards. Automated checks do not prove consent or clear intellectual-property rights. Screening logs contain decision reasons and request identifiers, not copies of raw prompts. Reported content is reviewed by the operator; requests or results may be restricted during an investigation. Review your full result before publication. A successful generation is not compliance approval.

Reporting and response targets

Use support@dolai.art or Contact. Include the page, asset/task reference and reason. For rights concerns, state the affected right and your relationship to it. Do not attach illegal imagery or unrelated personal information; provide its reference. Reports are restricted to people handling the case and reporters are not publicly identified.

Urgent child-safety, NCII and credible-threat reports receive priority. Our targets are initial response within 24 hours for urgent reports, 2 business days for ordinary reports and a decision within 5 business days once facts are available. We can restrict content while investigating and explain material delays. These are operational targets, not a claim of 24/7 staffing.

Enforcement and appeals

Depending on severity, we can block generation, warn users, remove content, limit functionality, suspend accounts or terminate access. Serious/repeated violations may cause immediate restrictions. Credible illegal activity may be reported to appropriate authorities when required or appropriate. Refunds and credits follow the Refund Policy; unrelated purchases are not automatically forfeited by a safety block.

Appeal through support with the task/decision reference and supporting facts within 14 days of notice. Our target is review within 10 business days. Relevant records may be preserved for investigation/legal duties. We update policies and safety controls as risks change.