About the AI Checker Q&A Forum
The AI Checker forum is a question and answer community for people who have to make sense of an AI detector score. A student whose own essay came back flagged. A teacher holding a report and wondering what to do with it. An editor who has to decide whether a freelancer's draft goes ahead. A writer learning English as a second language whose clear, careful prose keeps getting marked. You ask a specific question, and you get a specific answer that names the evidence, the limits and the next step.
It sits beside aichecker.tech, a free AI detector and plagiarism checker. The checker stores no submitted text, needs no account, and publishes how it works at /methodology. It says plainly that a detector score is a probability, not proof. The forum is where the harder questions land: what a number means, why two tools disagree, and what a fair process looks like when a score is the only thing on the table.
What we cover
There are twelve sections, each for questions in its own area.
- Reading detector scores. What a percentage means, what a confidence band is, and why one score from one tool is a weak basis for a decision.
- False positives and flagged work. Human writing that gets marked, why it happens, and how to respond calmly.
- Turnitin AI writing reports. How to read the report, what the highlighted passages do and do not show, and what to ask your institution.
- Comparing AI checkers. How to choose a tool, how to test one fairly, and what claims to be wary of.
- How AI detection works. Plain explanations of the methods behind detectors and their known weaknesses.
- Teachers and grading decisions. What teachers do with a result, and how to build a process that does not rest on a score alone.
- Students and appeals. How to prepare for a conversation about flagged work and how an appeal usually proceeds.
- Editors, clients and publishing. Handling detector results in a newsroom, an agency or a freelance relationship.
- ESL and second-language writers. Why some detectors misread non-native writing, and how to protect your work.
- Plagiarism and citation checks. Matching reports, citation style and the difference between copying and similarity.
- School and workplace AI policies. Writing, reading and questioning rules about AI use.
- Detector tests and experiments. Members and staff share how they tested a tool, with the method written down so others can repeat it.
Who answers
Early answers are written by the forum staff. Corinne Vale is the forum editor. She writes on reading reports, appeals, and classroom and newsroom policy. Anselm Okoro is the detection methods editor. He writes on how detectors work and how to test them. Staff posts carry a visible Staff label beside the name, so you can always tell when you are reading the team and not another member. The forum is run by the aichecker.tech team, so staff may refer to the checker where it is relevant, and they say so when they do.
Staff answers open with a short plain answer, then explain. Members are welcome to answer too, and the best answer in a thread is chosen on its merits, not on who wrote it.
How answers use sources
Answers cite sources. A claim about how a detector works, what a vendor says its tool can do, or what an institution requires should point to something you can check: a published methodology page, a vendor's own documentation, a peer reviewed paper, or an official policy. Where a source is a vendor describing its own product, the answer says so. Where we describe our own testing, we say what we tested and how.
We also try to be honest about uncertainty. Detectors are not settled science, and a good answer says what is known, what is disputed and what depends on your situation. If an answer sounds more certain than the evidence allows, say so in a reply.
Scores are probabilities, not proof
This is the idea the whole forum rests on. A detector returns an estimate of how likely it is that text resembles machine output. It does not identify an author, and it cannot show how a piece of writing was made. Answers here treat a score as one input to a human decision, never as a verdict. Nothing in this forum is a ruling on whether a particular person cheated.
Corrections
If an answer is wrong or out of date, tell us. When we correct a staff answer, we add a dated note in the thread saying what changed and why, so nobody has to guess what was edited. Tools and policies change often, so staff answers carry the date they were last checked. You can reach the team through the contact page, and the posting guidelines explain what the forum allows.
Who answers here
Corinne Vale
Forum editor, aichecker.tech
Corinne edits the aichecker.tech forum. She writes the answers on reading reports, appeals and classroom and newsroom policy, works from the published guidance of the tools and institutions involved, and says plainly when a score cannot settle a question.
Anselm Okoro
Detection methods editor
Anselm covers how AI text detectors work and how to test them. He reads the vendors' own documentation and the research behind the accuracy claims, and he marks clearly where the evidence is thin or the claim cannot be checked.