1 item
Instantly detect AI-generated text with inline highlighting. Works like spell-check for AI content. ## SHORT DESCRIPTION (132 chars max) Detect AI-generated text on any webpage with linguistic analysis. Works locally, zero tracking. --- ## FULL DESCRIPTION WHAT IT DOES TruthLens highlights potentially AI-generated text as you browse. Scan any webpage, document, or email to identify passages that show linguistic patterns common in AI-written content. It works like spell-check, but for content authenticity. HOW IT WORKS TruthLens analyzes text across six linguistic dimensions: - Perplexity: Consistency of word predictability - Burstiness: Variation in token probability distribution - Vocabulary Diversity: Range and repetition of word choice - AI Phrase Fingerprinting: Detection of common AI-generated patterns - Structural Analysis: Sentence complexity and coherence patterns - Repetition Mapping: Unnecessary word and concept redundancy Suspicious passages are highlighted inline: red squiggly for likely AI content, orange for uncertain cases. Everything is analyzed in your browser—no text is sent anywhere. YOUR PRIVACY, PROTECTED All analysis happens locally in your browser. We never collect, store, or transmit any of the text you scan. No accounts, no tracking, no data servers. Complete privacy by design. WHO SHOULD USE THIS - Educators screening student work - Journalists and content creators assessing sources - Publishers reviewing contributor submissions - SEO professionals monitoring content integrity - Anyone concerned about AI-written content in their feeds PRICING - Free: 5 scans per day - Pro: £19.99/month for unlimited scans IMPORTANT DISCLAIMER TruthLens is a first-pass screening tool that flags suspicious linguistic patterns. It is not a definitive detector of AI content. Like any automated analysis, it can produce both false positives and false negatives. Use it as part of your broader content evaluation process, not as absolute proof of AI generation.
rating_count is the Chrome Web Store ratings count, not a written-review count.