AI Content Policy

1. Inputs and permissions

• Upload or reference only materials you are authorized to process, respecting copyright, trademark, publicity, privacy, client permissions, and confidentiality.
• Do not submit client secrets, identity documents, health, financial, or other sensitive personal information that a job does not require.

2. Human review is required

AI output may contain errors involving facts, text, people, products, brands, cultural context, or visual details. A qualified person should review any external publication, client delivery, or commercial use.

3. Prohibited content and uses

• Illegal activity, fraud, impersonation, phishing, malware, or evasion of safeguards.
• Non-consensual intimate content, sexual exploitation, and any sexual content involving minors.
• Hate, harassment, threats, incitement of violence, or targeted degradation based on protected characteristics.
• Unauthorized deceptive impersonations of real people, identity documents, evidence, or political manipulation content.
• Relying on unverified AI output for medical, legal, financial, or other high-risk decisions.

4. Real people and synthetic media

When identifiable people are involved, obtain necessary permissions and do not imply statements, conduct, or endorsements they never made. Clearly disclose AI generation or modification when required by law, contract, industry rules, or platform policy.

5. Intellectual property and similarity

Do not ask models to reproduce protected works or brand assets for infringement. Outputs may resemble existing works; conduct appropriate copyright, trademark, and similarity review before important commercial releases.

6. Model safeguards

SKAND or upstream providers may refuse, modify, stop, or review requests that may violate these policies. Do not split prompts, change wording, or switch models to evade safeguards. Any unused Credits are handled under the Credits Rules.

7. Responsibility and reporting

Users and their teams are responsible for how generated content is ultimately used, modified, published, and distributed. If you identify suspected illegal activity, infringement, or a serious safety issue, stop using the content and contact your team admin.