AI Usage Policy

This policy describes how AI features in Imgraft work and how they may be used.

Independent interface and model providers

Imgraft is an independent interface. Image generation is provided through APIMart; image and text moderation is provided directly by OpenAI. Imgraft is not affiliated with or endorsed by the model developers.

Standard and HD are Imgraft service tiers, not model names. The change-strength selector adds prompt instructions; it does not set a numeric denoising parameter. Generated details can differ from your source even when you ask to preserve them.

What you may submit

Submit only content that you own or are authorized to use. Do not submit passwords, API keys, payment card data, government identifiers or other sensitive personal data unless a feature explicitly supports it.

How inputs are processed

Task inputs are sent to third-party AI model providers to generate results and are stored with the task for delivery, billing, support and abuse prevention. Providers process them under their own terms and do not receive your account credentials.

Generated results

AI output can be inaccurate, incomplete, biased or similar to content produced for others. You are responsible for reviewing results before use. Do not rely on them as the sole basis for medical, legal, financial, employment or other high-impact decisions.

Prohibited uses

Face swaps, deepfakes, face manipulation, NSFW, adult, explicit and sexually suggestive content are prohibited. Do not use the Service to change a person's identity, face or body. Edits that keep the person the same, such as changing the clothing or the background of an adult photo you own or have permission to use, are allowed under the Acceptable Use Policy.

You may not use AI features to create unlawful, harmful, deceptive or infringing content, to impersonate others, or to attempt to extract model or system internals.

Reporting

Report unsafe output, rights concerns or failures to support@imgraft.com and include the task ID.