Overview
TextInspector's AI Detector is a sophisticated forensic tool built upon the legacy of the original TextInspector linguistic platform developed by Professor Stephen Bax. In 2026, the tool has evolved into a multi-model ensemble system that distinguishes between human-authored text, AI-generated content (GPT-4o, Claude 3.5, Gemini Ultra), and hybrid 'cyborg' content. Unlike generic detectors, it leverages deep linguistic features including Lexical Diversity, CEFR alignment, and Part-of-Speech (POS) distribution to identify the 'statistical fingerprints' left by Large Language Models. Its architecture is specifically optimized for high-stakes environments like academic research and legal documentation, where false positives can have severe consequences. By combining traditional corpus linguistics with modern transformer-based classifiers, it provides a 'Probability of AI' score alongside a detailed breakdown of sentence-level perplexity. This hybrid approach allows users to see not just *if* a text is AI-generated, but *why* it deviates from natural human linguistic patterns, making it a preferred choice for university admissions and professional publishers.