AI Ethics & Safety

1. Our position

Generative video is a powerful tool and, like any tool, it can be misused. This policy states what Sharpy - operated by - does to keep the platform safe, what we expect from you, and how enforcement works. It applies to every prompt, upload and output on the Service, and forms part of the Terms of Service. The Service is intended for users aged 16 and over (or the minimum digital-consent age in your country, if higher); accounts of users below that age are closed when we become aware of them.

2. Prohibited content

You may not use the Service to generate, upload or distribute content that:

• sexually exploits anyone, or depicts minors in a sexual context - absolutely prohibited, reported to authorities, and grounds for immediate termination;
• depicts a real, identifiable person without their consent - including deepfakes, synthetic endorsements and impersonation of anyone who did not agree;
• incites hatred, harassment, discrimination or violence against individuals or groups;
• promotes terrorism, violent extremism, or detailed instructions for weapons, drugs or other serious harm;
• is deceptive in a way that could cause real-world harm - fabricated news footage of real events presented as authentic, election disinformation, fraud or impersonation;
• violates privacy - surveillance-style imagery of people in private settings, or content built from personal data you have no right to use;
• infringes intellectual property, per the Copyright & Licensing policy.

3. Transparency: marking synthetic content

Every video delivered by the Service carries a visible Sharpy watermark. It exists so that anyone who sees the video knows it was AI-generated. The watermark cannot be disabled, removed or obscured - attempts to do so violate the Terms and this policy. We chose a visible, pixel-level burn-in deliberately: it survives re-encoding and cropping better than metadata alone. We will re-evaluate forensic (invisible) watermarking as the technology matures.

This is our transparency commitment for synthetic content, aligned in spirit with Article 50 of the EU AI Act: material generated by the Service is marked in a machine-noticeable and human-visible way as artificially generated, so that synthetic media is distinguishable from authentic recordings. Where you republish an Output on another platform, you are responsible for any additional AI-disclosure labelling that platform or your local law requires.

4. What we do on our side

• Provider policies apply. Generations run through third-party model providers whose own safety systems and acceptable-use rules also filter requests; a prompt refused upstream is refused here.
• Accountability. Every generation is tied to an authenticated account, and administrative actions on the platform are audit-logged. Abuse is traceable, which is itself a deterrent.
• Private delivery. Uploads and outputs live on private storage; downloads use short-lived signed URLs, so there are no public permanent links to content we host for you.
• Curated showcase. The public showcase shows only hand-picked videos that comply with this policy.

5. Reporting and enforcement

1. Report abuse to [email protected] with as much detail as you can (what you saw, where, when). Reports are read by a human, not a bot.
2. Depending on severity, enforcement ranges from content removal and a warning, to suspension or termination of the account and forfeiture of remaining credits.
3. Illegal content - in particular anything involving minors - is reported to the competent authorities.
4. If we act against your account you will, where lawful and practicable, be told why and how to appeal by replying to the notification email.

6. Limits of this policy

No moderation approach is perfect: probabilistic generation means harmful output can never be reduced to zero, and filters can occasionally block legitimate creative work. We prefer the occasional false positive to the alternative. This policy evolves - the “Last updated” date above reflects the current version, and material changes are announced in the Service.