Safety

Acceptable Use Policy

Updated July 2026

Prohibited Uses

FreeArtAI must not be used to create, request, distribute, or facilitate:

• Real-person deepfakes, face swaps, non-consensual use of another person's likeness, or attempts to impersonate a real person or organization.
• Any sexualized content involving minors or content that exploits, endangers, or depicts the abuse of a minor.
• Pornographic content, extreme violence, gore, or graphic injury.
• Hate speech, targeted harassment, threats, or doxxing.
• Election manipulation, voter deception, or content that falsely presents itself as coming from a real person, public official, business, or institution.
• Content that infringes third-party copyrights or trademarks, or otherwise violates another party's intellectual-property rights.

How We Enforce This Policy

Every prompt is checked against a local keyword safety policy before the generation model is called. Requests that match prohibited categories are blocked and are not sent to the model.

Prompts that pass the local check are reviewed a second time by the FLUX built-in safety checker during generation. The safety checker remains enabled for all production requests.

Anonymous use is limited by hard quotas of 3 generations per IP per day, 2 attempts per IP per minute, and 500 generations site-wide per day. Accounts or access associated with repeated or serious violations may be suspended without a refund.

Reporting Violations

Report suspected abuse, non-consensual imagery, impersonation, or intellectual-property concerns to hello@freeartai.org.