Your AP Seminar paper comes back flagged, and the first thing you want to know is what AI detector does College Board use. It is not on your score report, so rumor fills the gap with guesses about Turnitin and Copyleaks.
College Board does not name the detector it runs on submissions, and a flag is not proof of anything. Knowing which tool produced it, and how these tools misread formal academic prose, changes how you respond to a flag. Run your draft through AI Busted first.
What AI Detector Does College Board Use on AP Work?
Start with College Board's own published policy, because College Board does not give a direct answer naming the detector it runs on AP Seminar papers. Its help center explains what happens once work is flagged for unacceptable AI use, and that documented process is the only part you can verify. College Board builds no detector of its own; the screening tools schools license, such as Turnitin's AI writing detection, come from outside companies.
If your teacher names a specific checker, that name is what matters for your resubmission.
What You'll Need Before You Check What AI Detector College Board Uses
Gather these first.
- Your flagged submission and the notice wording.
- College Board's published AI policy, which names no specific detector.
- A detector that publishes its model list.
- Your draft history with timestamps.
Version matters more than the name. If you are already flagged, ask which tool and model version ran, since how College Board screens submissions varies by school.
How to Trace Which AI Detector Flagged Your AP Work
Have the flag notice and the draft it names open before you start. Nothing published answers what AI detector does College Board use, so you assemble the answer from documents.
- Read College Board's published AI policy first: Open the AP Central guidance on artificial intelligence tools and read the section on unacceptable use. It describes the behavior College Board investigates and reserves the right to treat inappropriate generative AI use as an academic integrity violation. It does not name a detection product. That gap is the result: a published rule, not a named vendor.
- Check the exact wording of your flag notice: The phrasing tells you where the screen ran. Notices that say "AI screen" or "AI detection tool" usually come from software your school licenses, and teachers comparing notes about the AI screen flagging original student work describe the same pattern. Write the exact phrase down before you reply.
- Ask your AP teacher or coordinator which tool your school licenses: This is the only step that yields a product name. Ask whether the school runs Turnitin, Copyleaks, GPTZero, or something else, and whether the screen happens before or after the work reaches College Board. A named answer beats any forum thread.
- Run the same draft through those detectors yourself: Paste the exact flagged text into the tools you were told about, plus a second opinion such as the AI Busted detector. Detectors disagree, so one tool calling a paragraph AI-written while another calls it human is normal, and that tells you the flag is not a verdict. If you used a paraphraser, check whether Quillbot text gets flagged before assuming the rewrite cleaned anything up.
- Compare the flagged sentences against what detectors measure: Detectors score text on signals like perplexity, how surprising each word choice is to a language model, and burstiness, how much sentence length varies. Read what an AI detector actually looks for, then mark the sentences that read flat and uniform. Keep a dated revision record beside them, because what happens after College Board flags a submission depends on what you can show.
The most common mistake is assuming College Board runs one named detector on every AP submission. Its published policy is about unacceptable use, and the screen that flagged you usually ran at the school level, on a tool your teacher can name.
The College Board AI Flag Checklist: What to Verify
These checks turn a vague flag into a dated record you can quote back. Each one produces something concrete: a saved file, a quoted policy line, a detector result, or a date.
| Task | Why it matters | Done |
|---|---|---|
| Save the flag notice and the exact draft version it names, including the date stamp on the file | The notice wording decides which screening rule applies to your case | - |
| Quote the College Board AI policy sentence your case falls under, word for word | Published policy, not forum rumor, sets your appeal grounds | - |
| Run the draft through Turnitin and Copyleaks and write down both percentages | One score cannot tell you which system flagged you | - |
| Mark every sentence in the flagged passages that you wrote before any AI help | Shows the human-written share of the paper | - |
| Log the teacher's name, the notification date, and your response deadline | Response windows are short and dated | - |
| Ask your teacher which screening tool the school licenses for AP Seminar | Schools often screen before College Board reviews a paper | - |
Two detectors disagreeing on the same paragraph is useful, not confusing. Quote both results side by side in your written response instead of picking the lower number, and describe what you changed between drafts. A reviewer reading your timeline can see the revision work; a single percentage cannot show that.
College Board does not name its AI detector
College Board does not name the AI detector behind its integrity screen, and that detail matters less than students assume. A flag opens a review, not a verdict, and the options a flagged student has include responding with context before a score decision.
The usual move is to collect your drafts and version history. That advice changes when the flagged submission was collaborative: shared outlines and group notes mean your evidence points at other people's writing as much as your own.
Frequently asked questions about what AI detector College Board uses
Does College Board actually use AI detectors?
Yes. College Board's published AI guidance states it reserves the right to investigate submissions where there is evidence of inappropriate generative AI use, and the integrity screen is how that review starts.
What happens after College Board flags your work for AI?
You get notified, and AP Central's help center page on flagged work covers the options a student has at that point. Read it before you reply, because your response is the record that matters.
Can a false positive affect your AP score?
A flag triggers an integrity investigation, not an automatic score change. What decides the outcome is whether you can show dated drafts and revision history that predate the notice.
How accurate are AI detectors on AP Seminar papers?
Accuracy varies by tool and by how much text it gets. Turnitin publishes its own accuracy figures for specific document sets, and short or heavily edited passages are where false positives cluster.
Conclusion: what to do next after a College Board AI flag
Treat the flag as a documentation problem first. Keep the dated drafts, the assignment sheet, and the notice together, because that record is what an integrity review actually weighs.
Then, before you submit the next paper, run it through AI Busted. Its detector shows you which passages read as machine-written, and its humanizer rewrites them into an academic tone you can defend line by line.
Choose differently if your teacher already accepts your drafts and revision history as proof. In that case, keep your process and skip the rewrite.
