1. Prohibited content and conduct
- Child sexual abuse material, grooming, exploitation, or sexualized depictions of anyone under 18, including fictional, generated, or age-ambiguous subjects.
- Non-consensual intimate imagery, sexual deepfakes, voyeurism, sexual violence, or explicit pornography.
- Instructions, planning, or glorification intended to facilitate violent wrongdoing, terrorism, self-harm, human trafficking, or illegal weapons activity.
- Fraud, scams, deceptive impersonation, identity theft, evasion of law enforcement, or other illegal activity.
- Hateful abuse, targeted harassment, doxxing, or content that promotes violence or exclusion against a protected group.
- Material that infringes copyright, trademark, privacy, publicity, confidentiality, or other rights, including use of a person's likeness without required permission.
- Malware, credential theft, attempts to bypass safeguards, automated abuse, account resale, or interference with providers and the service.
2. Consent, likeness, and disclosure
You must have the rights and consent required for each person, voice, brand, work, and reference you submit or depict. Do not present generated media as authentic evidence or use it to mislead people about a real event, endorsement, or identity.
Use clear disclosure when an audience could reasonably mistake synthetic or materially altered media for real media, especially in news, politics, advertising, health, and finance.
3. Moderation and traceability
Requests may be screened by Soho Frame and provider moderation before or after submission. A request can be rejected, a result quarantined, or access restricted when a policy or provider rule may have been violated.
Each job is tied to an authenticated account, workspace, project ID, generation ID, input assets, model route, timestamps, credit entries, and a provider task ID when one is returned. These records help us investigate a report without exposing another person's data.
4. Enforcement
We may warn a user, block a request, remove access to an output, limit models, return or withhold credits as the payment policy requires, suspend or terminate an account, preserve evidence, or report conduct to providers or authorities. Severity, context, history, and immediate risk inform the response. Attempts to evade safeguards can lead to immediate suspension.
5. Reporting
Report suspected abuse to support@sohoframe.com. Include the public URL, project ID, generation ID, and a short description when safe. Do not email, re-upload, or redistribute illegal content. If someone is in immediate danger, contact local emergency services.