AI Detector & Blocker

Spot AI-generated images, video and text as you browse, then tag, blur or hide it. Private: checks run on your device.

AI Detector & Blocker checks the images, videos and articles you see for signs that they were made with AI, and marks them so you can tell at a glance. ✦ AI: strong evidence, such as Content Credentials (C2PA) or IPTC tags declaring AI generation, Stable Diffusion or ComfyUI settings in the file, or leftover chatbot phrases in text. ✧ AI?: weaker hints, such as a platform's "AI info" label, AI-art hashtags, alt text or captions that mention AI, or AI-style writing. You choose what happens to each: Tag it with a badge (the default), Blur it until you click Show, Hide it completely, or turn that verdict Off. With Hide selected, new images stay invisible until they've been checked, so you don't glimpse them first. One switch at the top of the popup (or Alt+Shift+A) turns the whole extension on or off, and you can switch it off for individual sites. Click any badge to see exactly why it was flagged. Nothing is a black box. VERIFIED CONTENT CREDENTIALS When an image carries Content Credentials (C2PA), its signature is checked on your device with the official C2PA software against the C2PA trust list. You'll see who signed it (for example "verified: signed by OpenAI" on ChatGPT images, or "signed by Google" on Gemini images), and forged or edited credentials are caught. OPTIONAL: ENHANCED TEXT DETECTION Download a small AI language model (about 138 MB, only after you confirm) that runs entirely on your device to spot AI-written text the built-in checks miss. Remove it any time. PRIVATE BY DESIGN • Checks run in your browser. Nothing is sent to the developer, ever. • No accounts, analytics or ads. • Optional "deep checks" with Hive, Sightengine, GPTZero or Claude use your own API key and send content only to the service you choose. MORE • Reads the AI labels, community notes and AI-art tags that popular social sites add, so the right post gets flagged. • Right-click any image, video, text selection or page to check it. • Strict mode catches more AI writing, at the cost of more false alarms. • Turn it off for any site with one click. • The toolbar icon shows how many items on the page were flagged. HONEST ABOUT LIMITS No tool can detect every piece of AI content. Most social networks strip the hidden data that proves an image is AI, and well-written AI text can read like a person's. A flag is a strong hint, not proof, and no flag doesn't mean something is real. If the extension gets something wrong, use "Wrong? Report it" in any panel.