AI Academic Integrity Suite

The suite scans academic writing to identify possible AI‑generated passages, factual errors, and citation problems. By analyzing perplexity, burstiness, vocabulary diversity and style patterns, it assigns scores for hallucination risk, fact accuracy and AI‑authorship probability, and it can break down results by section to highlight specific issues.

Researchers and students can paste full papers, abstracts or excerpts, select which checks to run, and receive color‑coded findings with severity scores and suggested improvements within seconds. The tool also offers a “humanize” function that rewrites AI‑generated text at varying levels of aggressiveness while preserving citations, technical terms and overall structure.

All processing occurs locally in the browser with the user’s API key, and no data is stored or shared, emphasizing privacy for academic users. The service is positioned as a free, experimental, all‑in‑one integrity scanner for research‑focused writing.