# TruthScan TruthScan is an enterprise AI detection platform that analyzes submitted images, PDFs, text, voice, and video for AI generation or manipulation. It helps fraud, trust and safety, compliance, education, media, and government teams review content with forensic evidence and workflow-ready results. - Findings24: - Site: - Preços: Freemium - Categorias: Detector de deepfakes com IA, Detector de Conteúdo IA ## O que é TruthScan? TruthScan is an enterprise AI detection platform for identifying AI-generated and manipulated content. Its documented capabilities cover image and PDF analysis, while the broader platform is described as supporting text, voice, and video detection. It is intended for workflows where submitted media or documents influence refunds, claims, approvals, listings, onboarding, or publication. The platform provides a detection verdict with probability or confidence information, plain-language reasoning, and visual evidence such as image heatmaps. Image analysis covers fully synthetic images, AI-assisted edits, traditional digital retouching, and face swaps, including localized changes to otherwise authentic files. PDF analysis supports review of receipts, invoices, bank statements, proof-of-address files, and other submitted records. Teams can use TruthScan in a browser dashboard, through a Chrome extension, or through its API. The site describes REST API access with webhooks and batch or real-time processing, so detection can be placed at upload and higher-risk submissions can be routed to human review. A free plan is available, while paid plans use usage-based pricing. ## O que TruthScan pode fazer? ### Image and manipulation detection Analyzes fully AI-generated images, AI-assisted edits, traditional retouching, and face swaps. Region-level analysis can surface a localized alteration instead of averaging it across an otherwise authentic image. ### PDF and document analysis Screens submitted PDFs such as receipts, invoices, bank statements, proof-of-address files, and other supporting documents for signs of fabrication or manipulation. ### Forensic result evidence Each result includes a verdict, probability or confidence information, detailed reasoning, and—where applicable—a heatmap showing regions associated with generation or manipulation. ### API and workflow delivery The platform can be connected through REST APIs with webhooks and supports batch or real-time processing. Teams can use it at the point of upload and route higher-risk content to review. ### Browser and Chrome workflows Users can analyze files in the browser without relying on the API, including individual uploads and ZIP batches. The Chrome extension checks images from within the browser. ### Retention and deployment options Images are retained by default. Zero data retention is available on Business and Enterprise plans, with Enterprise options including regional processing, custom integrations, and dedicated or on-premises deployment. ## Casos de uso ### Returns and delivery claims E-commerce, retail, and delivery teams can screen damage photos, wrong-order evidence, and proof-of-purchase documents before issuing refunds or credits, while allowing ordinary submissions to continue through the normal flow. ### Insurance and healthcare evidence Claims and digital-health teams can review damage photos, injury imagery, medical invoices, and related documents for AI generation or editing before adjusters or clinical programs rely on them. ### KYC and financial onboarding Banks and other financial services teams can add a forensic check for identity images, bank statements, proof-of-address files, and other documents that may pass basic format or OCR checks despite being fabricated. ### Marketplace listing integrity Marketplaces can inspect seller-submitted product images and listing materials for synthetic or manipulated content before publication, helping trust and safety teams investigate questionable listings. ### Expense and lending review Finance, travel-and-expense, and lending teams can analyze receipts, invoices, pay stubs, and income documents before approving reimbursements or credit decisions. ## Perguntas frequentes ### How can a team use TruthScan? TruthScan can be used in the browser dashboard, through its Chrome extension, or via API. The site describes REST API access with webhooks and batch or real-time processing; the API is optional for browser-based workflows. ### What does a detection result include? Results include a probability or confidence score, classification or verdict, detailed reasoning, and a heatmap for visualizing relevant image regions. The platform describes these outputs as suitable for review and audit workflows. ### What file workflow is supported in the browser? Users can upload individual files or ZIP archives. The pricing FAQ recommends ZIP batches of 500 images or fewer for the most reliable results; supported image formats listed there are JPG, PNG, JPEG, TIF, and WEBP, with a 10 MB per-file limit. ### Does TruthScan detect localized edits as well as fully synthetic images? Yes. The site says its image pipeline covers fully AI-generated images, AI-assisted edits, traditional digital retouching, and face swaps, and analyzes regions so partial changes can be surfaced. Very small edits may be easier to assess after cropping to the area in question and rescanning. ### How are submitted images handled? Images are retained by default. Zero data retention is available on Business and Enterprise plans; with ZDR, submissions and customer data are discarded after detection and are not used for training. Regional processing is available on Enterprise and above. ## Informações rápidas - **Product type**: Enterprise AI detection and fraud-prevention platform - **Documented detection**: AI-generated and manipulated images and PDFs - **Additional modalities described**: Text, voice, and video - **Access methods**: Browser dashboard, Chrome extension, and API - **API workflow**: REST API with webhooks; batch or real-time processing - **Pricing model**: Free tier plus usage-based monthly plans and Enterprise pricing