FamilyAtlasAsk for the specific concern, the school’s AI policy and a chance for the student to explain their work. Keep the original submission, genuine drafts, notes, sources and available revision history. Do not manufacture a writing history, run the essay through “humanizer” tools or assume an opposing detector score proves innocence.
A detector output is not a complete misconduct decision. Turnitin’s current guidance explicitly says its indicator should not be the sole basis for action or a definitive grading measure. That limitation does not establish what happened in this student’s case; the actual writing process and school rules still matter. Discuss any permitted or unpermitted assistance honestly, including editing tools. Seek a calm review with the teacher and follow the school’s escalation process if needed. FamilyAtlas cannot determine authorship, certify an essay or overturn a school’s decision.
Turnitin: AI writing detection capabilities FAQs · Liang et al. / Stanford researchers: GPT detectors are biased against non-native English writers
The parent concern
An accusation can carry serious school consequences. Ask for evidence and process while preserving genuine work and avoiding claims that a detector proves either side.
Ask what evidence and rule are being used
Request the assignment instructions, AI policy in force at submission, disputed passages and the basis of the concern. Was it a detector, a source problem, a style change or something else? Ask how the student can respond and what the review timeline is. Be specific and respectful rather than accusing the teacher of bad faith.
Treat detector evidence with the right limits
Turnitin says it provides information for educators, not a determination of misconduct, and warns against using the indicator alone. A 2023 study by Liang and colleagues found bias against non-native English writing in the detectors it tested. That is historical evidence of a limitation, not a current error rate for every 2026 tool or proof about this essay.
Turnitin: AI writing detection capabilities FAQs · Liang et al. / Stanford researchers: GPT detectors are biased against non-native English writers
Keep genuine drafts and version history
Gather outlines, reading notes, drafts, citations and feedback that already exist. Google Docs documents viewing earlier versions from Last edit with edit permission; revisions can be grouped/merged and access varies. Show the actual available history, not a recreated timeline. For Word or another editor, check its own version/history support and storage setup rather than assuming all files retain it.
Let the student explain their choices
Prepare a short walkthrough of the argument, sources and revisions. Disclose assistance honestly and compare it with the assignment’s rule. Ask whether a discussion, review of drafts or supervised explanation can help establish understanding; the school decides its procedure. Do not coach the student to conceal real AI use or alter the original evidence.
Discuss schoolwork and ChatGPT use before the next assignment
Escalate through the school, then improve the process
If unresolved, use the school’s stated review/appeal route with the evidence and requested remedy. Ask for clear future instructions on AI, translation, grammar help and documentation of drafts. Local rules and accommodations vary; seek appropriate education advice for a serious unresolved penalty. No general article can promise an appeal outcome.
Respond if schoolwork prompts also disclosed personal information
What you can do right now
- Keep the original work and existing drafts.
- Request the concern, policy and review procedure.
- Explain the real writing process honestly.
- Discuss detector limits without treating them as proof.
- Follow the school’s escalation route if unresolved.
A way to start the conversation
“Let’s understand the exact concern and your actual process. We can ask for a fair review without inventing evidence or hiding help you used.”
A suggested script to adapt, not a clinical intervention.
What technology cannot solve
Neither a detector score nor a parent’s confidence proves authorship. FamilyAtlas does not administer school assessment, detect plagiarism or certify work.
Questions parents ask
Does a high detector percentage prove cheating?
It is not a complete decision. Turnitin warns against sole reliance; review the actual work and school policy.
Should we create new drafts to prove the process?
No. Keep genuine evidence and explain any gaps honestly; do not manufacture a history.
Does the 2023 study give the current detector error rate?
No. It concerns the tools and writing samples studied then, not every current model or this student’s case.
Sources and review
Provider documentation explains settings and eligibility. Professional guidance provides context, while our suggested family strategies are labeled separately. FamilyAtlas claims were checked against implementation, with release limits stated above.
- Turnitin — AI writing detection capabilities FAQs fact · checked 2026-09-30
- Liang et al. / Stanford researchers — GPT detectors are biased against non-native English writers guidance · checked 2026-09-30
- Google — Find what has changed in a file fact · checked 2026-09-30
Next source review: 2026-12-30. Menus, eligibility and rollouts can change; follow the linked provider instructions if your screen differs.
Educational guidance, not medical advice. The cited professional guidance is separate from our suggested family strategies. This article has not had independent clinical review; seek qualified help for persistent concerns or urgent safety risks.
Written by FamilyAtlas; product descriptions are not independent product recommendations. No claim of medical or psychological review is made.