Decopy for Web

Decopy — Content Authenticity Checker

Decopy is a detection service that assesses whether a piece of writing was produced by a person or by generative models (examples include ChatGPT, Claude, and Gemini). It separates original human prose from machine-generated or heavily paraphrased material and flags sentences that appear to be created by AI, helping maintain the trustworthiness of written content.

What it can do

• Highlights individual sentences that appear likely to be machine-generated, giving a visual map of where AI influence is suspected.
• Distinguishes straightforward AI output from content that has been edited or paraphrased to look human.
• Delivers a confidence metric that quantifies the probability of AI involvement in the text.
• Operates across multiple languages so you can evaluate content in a variety of linguistic contexts.
• Emphasizes accuracy to reduce incorrect attributions and improve reliability for sensitive checks.
• Offers in-browser analysis without requiring users to register an account.
• Is available at no cost for basic checks while enforcing strict confidentiality for submitted material.
• Serves practical needs in education, newsrooms, and corporate settings that need to verify authorship authenticity.

Access, cost, and privacy

Decopy is designed for quick, browser-based use. You can paste text and run an analysis without creating an account. Basic scanning is provided free of charge, and the system follows privacy safeguards to protect any text you submit from being shared or used beyond the analysis.

Who benefits from using it

• Educators validating student work or checking for AI-assisted submissions.
• Journalists and editors verifying the originality of articles and reports.
• Businesses enforcing content policies, ensuring compliance, and safeguarding brand voice.
• Researchers and institutions that need to check the provenance of written material.

How results are presented

Analyses return a clear, visual report: flagged sentences, an overall confidence score, and a concise summary of findings to help you interpret whether a piece was likely authored by a human or generated (or heavily altered) by AI. This makes it easier to act on the result—whether that means further review, revision, or documentation for compliance.