College Board AI Detector: The One Place College Board Says It Runs Turnitin on Your AP Work

College Board uses Turnitin in one published place: the AP Digital Portfolio, on final submissions for AP Seminar and AP Research (the two AP Capstone courses). The College Board pages we read name no AI detector for AP exam essays or for AP Classroom.

The source is College Board’s AP Capstone policy page, which states: “College Board uses Turnitin to evaluate all final submissions in the AP Digital Portfolio.” (AP Central, AP Capstone policies). The same page gives two purposes for the check: spotting plagiarism and spotting improper use of generative AI.

That same page, a few lines later, adds a warning in College Board’s own words: “Turnitin is a valuable tool for teaching students about citation, but it’s not a plagiarism detector. Feedback from Turnitin should be considered in context and not simply taken at face value.” Read the two statements together. College Board runs Turnitin on every final Capstone submission, and in the same breath tells teachers and students that its output is a prompt for human judgment, not a verdict. If you want to understand what that output looks like, our guide to Turnitin’s AI indicator and how it reads a paper covers the mechanics.

Only the AP Digital Portfolio has a named AI check, and it applies to AP Capstone final submissions. AP Classroom and AP exam day have no detector vendor on the College Board pages we read. The table sorts each AP product by what those pages say.

Sources: AP Capstone policies, AP AI-tools policy, About AP Scores. The practical upshot: College Board has published no detector for a single AP Classroom essay or an in-class assignment, so if one of those was flagged, the tool most likely came from your teacher or your school.

College Board has not published an AI-detection feature for AP Classroom. Its only named AI check, Turnitin, sits in the AP Digital Portfolio for AP Seminar and AP Research final submissions, which is a separate system from AP Classroom.

The two names get blurred constantly, so the distinction is worth spelling out. AP Classroom is where teachers assign practice questions, unit reviews and progress checks through the year. The AP Digital Portfolio is where Capstone students upload the performance tasks that count toward their AP score, and it is the system the Turnitin statement refers to.

Because College Board has published no such feature, a flag on AP Classroom work most likely came from your teacher’s own reading or from a detector the school licenses. Ask which one it was; the answer changes how the result should be read.

College Board’s scoring page describes AP free-response answers and through-course performance tasks as scored by people: college professors and experienced AP teachers at the annual AP Reading in June. That page does not describe AI grading of essays at all.

The wording, from College Board’s student-facing scores page: “The free-response section (essays and open-ended questions) and through-course performance tasks are scored at the annual AP Reading held during the first 2 weeks in June. Specially appointed college professors and experienced AP teachers score this section of the exam.” (About AP Scores).

Multiple-choice sections are machine-scanned against an answer key, which is ordinary answer-sheet scoring and a different thing from software judging your prose. Keep the two questions apart as well: who scores an essay (AP Readers) and whether a submission is checked for AI (Turnitin, on Capstone portfolio submissions) are separate steps run for separate reasons.

College Board requires every AP Seminar and AP Research performance task to be the student’s own work, checked through mandatory checkpoints with the teacher. Generative AI is allowed only as an optional aid for topic exploration, source searches, comprehension checks and grammar or tone, and skipping a checkpoint scores the task 0.

The checkpoint rule, from College Board’s help center: “During the performance tasks, students are required to successfully complete checkpoints with their teacher in order to demonstrate they have not used AI to bypass the work. Failure to complete the checkpoints will result in a score of 0 for that particular component of the AP Seminar and/or AP Research performance task.” (AP Central help center). Note what that means: the 0 for a missed checkpoint applies regardless of any AI finding. Checkpoints are the heart of the system, and Turnitin sits beside them.

Under its AP Capstone entry, College Board’s AI-tools policy lists the uses it treats as acceptable optional help: “Students can use generative AI tools as optional aids for exploration of potential topics of inquiry, initial searches for sources of information, confirming their understanding of a complex text, or checking their writing for grammar and tone.” (AP AI-tools policy). Drafting the argument itself is not on that list.

Rules differ by course, and the same policy page sets them one course at a time. AP Art and Design is the clearest: “Generative AI is allowed in the early stages of the artistic process. Using generative AI tools to create final submitted artwork is prohibited.” Art and Design students must also disclose any AI use in the Digital Portfolio for both Sustained Investigation and Selected Works. The page carries separate entries for other courses, including AP African American Studies, AP World Languages and AP Computer Science Principles; read your own course’s entry on that page rather than assuming one rule covers all of AP.

College Board does not treat a Turnitin result as proof either way: its Capstone policy page says the feedback “should be considered in context and not simply taken at face value.” One Virginia school district went further and ruled such detectors out as a sole source of detection. A score opens a human review; your checkpoints and drafting record settle it.

One high school’s own newspaper shows how that review can be triggered in practice. Lightridge News, the student paper at Lightridge High School in Loudoun County, Virginia, reported in June 2025 on AP Seminar and AP Research AI cases at the school, where roughly 15 to 20 students were caught in March and April 2025. A staff member quoted in the article described the escalation rule: “We usually use 50% as a threshold. If we give it to TurnItIn and it comes back 50%, we then go ahead and alert College Board.” (Lightridge News, student newspaper of Lightridge High School, June 11, 2025).

The same article puts the district’s position plainly: “These detectors are so unreliable as to have been recently deemed by LCPS as unusable as a sole source of detection. A paper reported by a detector as 52% AI one day might come back as 43% the next.” That 52% and 43% pair is the article’s general example of how detector scores move; it is not a report about any specific escalated paper. It lands on the same point College Board makes about Turnitin: a single number opens the question, and the student’s documented process answers it.

If English is your second language and your writing is formal and tightly edited, our page on AI flags and non-native English writing explains the research on why such prose gets a second look and what helps.

A student whose Capstone work is flagged, whether an IWA, an IRR or an AP Research paper, gets three options from College Board: take a 0 on that component and keep an AP score, accept cancellation of the whole exam, or request reconsideration. A flag opens a decision; it is not an automatic penalty.

College Board’s help center sets out the three options in these words (AP Central help center):

1. “Voluntarily accept a score of 0 on that particular component of the performance task, and still receive an overall AP score.”
2. “Voluntarily accept cancellation of the entire exam.”
3. “Request a reconsideration of that particular component of the performance task.”

If reconsideration is denied, a student can still ask for score cancellation by replying to the decision email within 3 business days (score cancellation after reconsideration).

A reconsideration request rests on three kinds of evidence College Board names: version history (such as a document’s dated edit history), a teacher attestation of the progress they watched, and checkpoint documentation or notes. Our step-by-step guide to showing your own authorship with your drafts and edit history walks through gathering them.

Yes. You can run your IWA, IRR or research paper draft through our free AI detector before it goes into the AP Digital Portfolio, with no account for short checks. The scan goes to no school or College Board system, and we count it without saving your text.

Here is exactly what the free detector gives you:

• Without an account: 4 scans a day, up to 250 words per scan. That is enough to test the sections that worry you most, such as an introduction or a dense line of reasoning.
• Signed in: a single scan takes up to 60,000 characters, so a long research paper fits in one pass.
• Verdicts in plain words: Human, Inconclusive, AI Detected, or Too short. Passages under 60 words return Too short rather than a guess.

Use the result the way College Board asks everyone to use Turnitin’s: as information. A flag or an Inconclusive read on a draft you wrote is a prompt to put your drafts and checkpoint notes in order, not a verdict on your work, because those records are what show the paper is yours. Run your draft through the detector now, then keep writing in the document that holds your version history.