See what the labels mean, decide what to review, and make your writing clearer.
Your result is an estimate based on writing patterns. Read the overall label, percentage breakdown and highlights together.
The sentence at the top summarises our assessment of your text, based on the balance of Human, Assisted and AI content and the confidence of the detection.
The conclusion describes the text overall. The percentage breakdown and highlights show the details.
The main percentage combines the portions of your text marked AI and Assisted. It does not tell you the probability that the author used AI.
The remaining 75% is classified as Human.
| Category | What it means |
|---|---|
| Human | Classified as human-written. No meaningful AI involvement detected in this stretch. |
| Assisted | Shows characteristics of both human and AI writing. Most often this means human-written text that was run through an AI tool to polish, rephrase, or extend it, or an AI draft that was substantially hand-edited afterward. |
| AI | Classified as likely AI-generated. Reads as generated by an AI model with no meaningful human authorship. |
These categories are statistical estimates, not certainties and not legal proof. Highlights do not establish how the text was written. Treat every result as one input among several rather than the sole basis for a decision.
Confidence describes how clear-cut the scoring was. It is separate from how much AI is in it.
The location description, highlights and "Start of text" to "End of text" strip show where flagged passages appear.
Focus on making the ideas and wording your own. These edits can improve your writing, but cannot guarantee a lower score.
Write down what you mean, why it matters, and what supports it.
Set the existing text aside and explain the idea in your own words.
Use relevant examples, observations, or evidence you can support.
Explain why your evidence matters and how you reached your conclusion.
Remove repetition and filler. Keep a tone that suits your audience.
Fix awkward wording, then compare the same complete passage.
Illustrative examples. Use details only when they are true for you.
| General wording | More specific wording |
|---|---|
| Effective communication plays a crucial role in improving team productivity. | Our handovers work better when each task has a named owner and a deadline. |
| This innovative solution enhances efficiency and optimizes productivity. | The tool copies order details into our planning sheet, so the team enters the information once. |
| The project provided valuable learning opportunities. | Two suppliers cancelled during the project. Next time, I would confirm a backup supplier before setting the delivery date. |
These are writing examples, not detector test results.
Detectors can make mistakes. Before rewriting, check your text and keep evidence of your writing process.
Still think the result is wrong? Email support@werida.io with a link to the result - we do look at these individually.
AI Checker analyses patterns in word choice, sentence structure and how ideas connect. It uses these patterns to assess whether writing resembles AI-generated text.
Your result includes an overall label, an AI involvement percentage, a category breakdown, confidence and highlighted passages.
Open your individual report from History. Read highlighted sentences with the surrounding text. Results are estimates and can flag text you wrote yourself.
"Human written" is the overall assessment of your text. Some passages may still be classified as Assisted and highlighted in yellow. Read the percentage breakdown and highlights alongside the overall label.
AI Checker analyses patterns across passages of text. Longer passages generally provide more context and support a more reliable assessment than a single sentence.
The surrounding sentences may influence how a sentence is assessed. When you add, remove or rewrite nearby text, you change that context. This may explain why a sentence receives a different highlight after a new check, even when its wording stays exactly the same.
The percentages are also recalculated for the updated text. Removing one highlighted passage can therefore change both the overall breakdown and the highlights elsewhere. Recheck the complete passage and review highlighted sentences alongside the surrounding text.
Rewriting can change the patterns the detector analyses. The result alone cannot establish which tool was used or distinguish a grammar correction from substantial AI rewriting.
Not necessarily. Detectors use different models and scoring methods, so their percentages may mean different things. Compare the explanations and highlighted passages as well as the headline score.
There is no guarantee. Rewritten text can still be flagged. Revise from your own notes, add accurate details and check that any edits preserve your meaning.
No. This check estimates AI involvement in your writing. Plagiarism checking compares text with existing sources. Review quotations and references separately.
Email support@werida.io with a link to the result (use the Share result button) and we'll take a look.
For everything else, see the main Support page.