Acceptable Use Policy
Content standards, six prohibited AI generation categories, document editing restrictions, and enforcement rules for TwinFont.
Última actualización: 2026-11-16
TwinFont edits and repairs text in images using in-browser tools and optional cloud AI background repair. Because cloud AI image repair generates rebuilt pixels, clear content boundaries apply to every input and output processed through the Service. By using TwinFont, you agree to the rules below.
What TwinFont is for
- Fixing typos, dates, prices, and names in your own graphics, posters, banners, and social posts.
- Localizing or translating text in marketing assets you have the right to modify.
- Repairing garbled text in AI-generated images.
- Building mockups, demos, and documentation from product screenshots.
- Redacting personal information (names, emails, phone numbers) before sharing a screenshot.
- Removing captions, timestamps, or labels from images you own or are licensed to alter.
Content Standards & Prohibited Categories
TwinFont strictly prohibits generating, editing, or facilitating any of the following six categories of prohibited AI-generated or edited content:
- Pornography / NSFW content — sexually explicit, adult, or non-consensual intimate imagery.
- Violence / gore — graphic violence, blood, gore, extreme physical harm, self-harm, or content promoting terrorism or mass violence.
- Hate speech & discrimination — content that promotes hatred, dehumanization, harassment, or discrimination against individuals or groups based on race, ethnicity, nationality, religion, gender, sexual orientation, disability, or other protected characteristics.
- Child-unsafe content (CSAM) — child sexual abuse material or any sexual, exploitative, or harmful content involving minors (zero tolerance; reported to relevant authorities).
- Deepfakes / impersonation of real people — face swaps, nudification, fabricated statements, or misleading AI edits depicting or impersonating real individuals without lawful basis, or deceptive impersonation of brands and organizations.
- Copyright, trademark & IP infringement — unauthorized editing, reproduction, or watermark removal involving third-party copyrighted works, trademarks, brand assets, or publicity rights.
Prohibited Document & Deceptive Edits
In addition, you may never use TwinFont to create or alter an image to deceive someone, forge records, or misrepresent a fact. In particular, you may not edit:
- Identity documents — passports, national ID cards, driver licences, residence permits, visas.
- Financial instruments and records — bank cards, cheques, bank statements, balance or transfer screenshots, payment confirmations.
- Receipts, invoices, and order confirmations — including delivery and tracking screenshots.
- Certificates and official documents — diplomas, licences, test results, medical records, legal notices, government letters.
- Chat and message screenshots intended to fabricate what someone said.
- News, quotes, or statements attributed to real people or organizations in a misleading way.
- Other people's copyrighted work or watermarks without permission, including removing a creator's watermark to reuse their work.
- Trademarks and brand assets you do not own or are not licensed to use.
- Source of uploaded photos: You may only upload images containing real people if the photo is your own or you have the depicted person's permission, and never for deepfakes or unlawful edits.
Content Moderation & How We Enforce This
- Automated screening & provider filters: Before cloud AI background repair runs, TwinFont runs cloud vision classification to flag and block restricted document types (
id_document,bank_card,receipt_invoice,certificate,payment_screenshot), and upstream AI model provider safety filters screen requests and outputs against CSAM, sexual content, violence/gore, and hate speech. Automated detection is a technical safeguard, not a complete verification or permission — being technically able to edit an image does not make it permitted. Local blur/redaction remains available solely for legitimate privacy protection. - Model disclosure & provenance: Cloud background repairs are rendered using third-party AI models (Google Gemini models via APIMart), which may embed provenance signals in generated outputs.
- Account enforcement: Accounts used for prohibited edits or repeated policy violations are suspended or permanently terminated immediately without refund, and we cooperate with lawful requests from authorities.
Reporting Misuse & Abuse
If you believe an image edited or repaired with TwinFont violates these content standards or is being used to deceive, please email support@twinfont.com (or use the feedback/support link in the footer) with the image, URL, or details of where it appears. We review abuse reports promptly (acknowledging critical safety reports as soon as possible and responding within 2 business days).
Changes
We may update this policy as the product and compliance requirements evolve. Material changes will be published on this page with an updated date above.