Responsible-use library

    Practical guidance for reading automated results carefully

    Use these guides to interpret AI signals, review source matches, investigate fairly and keep context in every consequential decision.

    Start with score interpretationFor educators

    Eight focused guides

    Choose the question you need to answer next

    The library avoids detector folklore and magic thresholds. Each guide moves from an automated output to evidence a person can inspect.

    A practical route

    Move from an automated signal to a reviewable decision

    Choose the guide that matches the output in front of you, then widen the review before deciding what it means.

    1. 01

      Identify what the tool actually reported

      For an AI-writing percentage, begin with score interpretation. For matched passages or sources, begin with the difference between similarity and plagiarism. This keeps two different signals from being treated as the same claim.

      Interpret an AI score
    2. 02

      Test plausible alternatives

      Check whether sample length, writing style, editing, document preparation or an ordinary source match could explain the result. A useful review looks for evidence that challenges the initial concern as well as evidence that supports it.

      Review false positives
    3. 03

      Place the result inside a fair process

      Use drafts, notes, version history, cited sources, the applicable policy and the writer’s explanation. Record the evidence considered and the reason for the final human judgement instead of using a detector threshold as an automatic verdict.

      Follow the review workflow

    Shared foundation

    Every guide begins with the same limitation

    A detector analyses the submitted text. It does not observe the writing process or know the rule that applies.

    AI-writing signals can be useful for triage. Source matches can focus a citation review. Neither output establishes authorship, intent or academic misconduct without additional evidence and a fair process.

    Product context

    Know the method before relying on the output

    Review how Genutext converts scores, handles sentence results, separates source candidates and limits data retention.

    Read the methodologyHow it works