Which Universities Use Pangram? A 2026 Check of School AI-Detector Policies

Table of Contents

• The short answer, as of the last-checked date

• The schools, one by one

• UT Austin: what its page actually says

• Mentioned with Pangram, but not adopters

• How many schools use any detector at all

• The policy terms to look for

• How to check your own school

• If your work is flagged

• FAQ

Last checked: October 27, 2026. These policies change by term, and this page will be rechecked each academic term.

The short answer, as of the last-checked date

Among the school pages we read, the University of Chicago is the clearest case: Pangram is available to all UChicago instructors starting in the Autumn 2026 quarter. Dartmouth's Faculty of Arts and Sciences has authorized Pangram as its only approved detector, though the college hasn't bought an institutional subscription. Wellesley piloted it in late 2025 and has since opened it to any faculty member who wants it, and Portland Community College is evaluating it with limited licenses.

One caution. "Available" or "authorized" means an instructor may use Pangram. It doesn't mean every paper you turn in gets scanned.

The schools, one by one

Each entry says what the school's own page, or credible reporting about it, says, and stops there.

University of Chicago: available to all instructors from Autumn 2026

Booth Instructional Design, the business school's teaching-support team, has the plainest statement. Its Pangram page opens: "Newly available to all UChicago instructors in the Autumn 2026 quarter, Pangram is an academic integrity application that specializes in AI writing detection." Pangram is "fully integrated into Canvas" and can be turned on for assignments students complete on their own; it "is not currently available for group assignments." Instructors can also use a web dashboard or a Chrome extension. Results come back as the likelihood a submission is "AI-written," "AI-assisted," or human-written. The page shows no date; we read it October 27, 2026.

How the page tells instructors to use it

For a rollout page, it's careful. It recommends "approaching AI detection tools cautiously – as imperfect resources for engaging with writing in the age of GenAI rather than objective assessors," and adds, "We recommend being upfront about your use of AI detection tools, both for reasons of transparency and equity." It asks instructors to decide in advance what follows a flag: "You might consider having a conversation with the student during office hours or reaching out via email to inquire about how they produced their writing."

What the central teaching page adds

Booth is one school describing a university-wide rollout, so it's fair to ask whether the center agrees. It does. The Chicago Center for Teaching and Learning's Navigating Academic Integrity in the AI Landscape says Pangram "is available to those who teach at the University, including the College," and that "particular schools, divisions, and programs may have specific policies or guidance on whether or how to use Pangram." It tells instructors to "clearly state if you plan to use the Pangram AI detection tool provided by the University," and puts the limit plainly: "A Pangram result, in isolation, is not enough to conclude a student made inappropriate use of AI." It also notes that "not all instructors will choose to use Pangram." One line matters to students: once Pangram is on for a Canvas assignment, "all submissions will be scanned," not only the suspicious ones. UChicago IT's Pangram-Canvas article adds that "the University has an enterprise license for Pangram." Neither page shows a date; both were read October 27, 2026.

Dartmouth: authorized for faculty, no institutional subscription yet

Dartmouth's position sits in a short PDF, the Faculty of Arts and Sciences Guidelines on Use of AI Detection in Teaching, marked "Adopted September 2026" and read October 27, 2026. The student paper covered it in "Dean of faculty authorizes professors to use AI detector Pangram to review student work," by Jackson Hyde, published October 1, 2026. Interim dean of the faculty John Carey sent the guidelines to undergraduate faculty on September 29, the paper reported.

What the guidelines say

There are seven. Faculty may submit student writing to AI detection software. "As of September 2026, Pangram is authorized for this purpose because it has shown the most reliability of available detectors." Faculty must tell students in writing whether work may be submitted and whether results could affect grades. They must remove all personally identifying information first. They may not use a result to affect a grade "without first communicating with the student about the quality and integrity of their work." The Committee on Standards "looks for a constellation of data (that is, information beyond the result of an AI-detection tool)" before finding a student responsible. And Dartmouth "is evaluating whether to purchase an institutional subscription to AI detection software, but no final decision on that has been made." Only the instructor of record may submit student work.

Two details students won't expect

The privacy rule goes past blacking out a name. A footnote says it covers "identifying metadata embedded in the file itself, such as author or reviewer names stored in the document properties or commenting features of Word, PDF, and similar files." Personal narratives can't be submitted "unless that information has been removed or modified so that the student cannot reasonably be identified." The essay about your grandmother's kitchen stays out unless it's been altered.

And it's optional. Carey told The Dartmouth that faculty "are not required" to use AI detectors.

These are Arts and Sciences guidelines, sent to undergraduate faculty, and nothing we read extends them to Dartmouth's professional schools. HumanizerNerd, which tracks school detector policies, issued an October 23 correction listing Dartmouth as "Allowed, with cautions," not "Provided." That's the right way to say it.

Wellesley College: a fall 2025 pilot, then opened to faculty

The Wellesley News, the student paper, reported in "Wellesley pilots AI checker Pangram amid faculty concerns about AI" (Jessica Chen and Noufeesa Yahyaoui, December 10, 2025) that Library & Technology Services recommended Pangram "in late November" and invited faculty into a pilot that semester. That's student reporting, now ten months old.

There's a later word from inside the college. In a March 25, 2026 post on his Wellesley-hosted blog, Chief Technology Officer Ravi Ravishanker wrote that Pangram "was piloted by several faculty towards the end of the fall semester and based on positive feedback, we have now made it available to any faculty member who wants to use it." It's a personal blog, not a policy page, and we found no formal LTS page on Pangram's current status.

Portland Community College: a limited evaluation

PCC's Instructional Support Pangram AI Detection page says Pangram "is not perfect, and is being evaluated to see if it can support instructors in the short term while disciplines adapt to new approaches to learning assessment." The evaluation uses the Chrome extension, and access "is currently limited to faculty, staff, and students who are helping evaluate the platform." The page shows no date; on October 27, 2026, it still described an evaluation.

It also offers instructors sample syllabus wording: "This term, I will be evaluating Pangram's AI detection tool if I suspect any unapproved use of generative AI tools in student work." If you're at PCC, a sentence like that in your syllabus is the clue. Note the "if I suspect."

For context, a January 30, 2026 roundtable summary on PCC's teaching-center site says the discussion "highlights why PCC avoids predictive success analytics, automated plagiarism detectors, and digital proctoring tools due to built-in algorithmic biases." That's a college testing a tool, not handing it out.

UT Austin: what its page actually says

UT Austin's provost page on AI detection doesn't mention Pangram at all. What it does say is narrower than people tend to assume.

The page is AI Detection Software Guidance, from the Office of the Executive Vice President and Provost. It shows no date; we read it October 27, 2026. "The University does not endorse the use of AI Detection Software and does not have any central contracts or purchase orders with active AI Detection Software features." And: "The University prohibits the use of all third-party software, including AI Detection Software (ADS, software used to flag or otherwise detect output as AI-generated), to evaluate student work or assignments unless a University contract or purchase order is in place."

What that does and doesn't mean

The same section says contracts "must be administered by a department, College, School, Unit or at the enterprise level." So "no central contract" is the accurate statement. The page doesn't say UT bans Pangram, and it doesn't say whether any department holds a contract. It does close the side doors: detectors are "High Risk Software" that "cannot be purchased with a Purchase Card (Procard) or a personal credit card," and instructors who buy one that way "may be personally liable for paying any damages or legal costs."

Mentioned with Pangram, but not adopters

Some stories read like adoption claims and aren't. The University of Amsterdam's news outlet, Folia, ran the literature-review chapters of 50 public bachelor's theses through Pangram for an October 9, 2026 analysis. That's journalism, not a university decision, so UvA isn't a Pangram school on this evidence.

A Conversation piece by Mark A. Bassett, republished by Social Science Space on October 6, 2026, says Melbourne, Sydney, and UNSW "use AI-detection tools in some form," while Curtin and Queensland "have opted to not use AI-detection tools at all." It mentions Pangram only as a company selling detectors, so none of those schools can be tied to Pangram from it.

Pangram's own educator page says its models are "independently validated by researchers at the University of Maryland and University of Chicago." That's research, not a campus license, and it's the vendor talking. Its knowledge hub, last reviewed July 24, 2026, lists Google Classroom, Canvas, Schoology, Moodle, Blackboard, and Brightspace. A school running Canvas doesn't thereby have Pangram.

How many schools use any detector at all

HumanizerNerd, a site that covers humanizers and so sits close to this market, read 300 universities' own pages on October 9, 2026 and updated its count October 23. Of those 300, 53 provide a detector: 44 Turnitin, 4 Copyleaks, 1 Pangram (the University of Chicago), and 4 that don't name the product. 100 don't run one. Those are HumanizerNerd's counts from its own sample of large, well-known schools, not a census. If yours uses Turnitin instead, see our comparison of Turnitin, Originality, and GPTZero.

The policy terms to look for

• Authorized or approved tool. Which detector, if any, instructors may use (Dartmouth).

• Written notice. Whether instructors must tell you in the syllabus or in writing (Dartmouth requires it; UChicago recommends it).

• De-identification. Whether your name and file metadata come out first (Dartmouth).

• A conversation before grading. Whether you're contacted before a result affects your grade (Dartmouth requires it; UChicago suggests it).

• More than one piece of evidence. Whether a finding needs more than a score (Dartmouth's "constellation of data").

• Contract rule. Whether instructors may use only software the school has a contract for (UT Austin).

• Pilot or evaluation. Limited access while the school decides (Wellesley's pilot, PCC's evaluation).

How to check your own school

1. Search your school's site for "AI detection" on teaching-center, IT, academic-integrity, or provost pages. HumanizerNerd's method used the same kinds of pages.

2. Read your syllabus for detector language, like PCC's sample statement.

3. Ask your instructor. At Dartmouth and UChicago, the choice to use a detector sits with them.

4. Write down the page and the date, because these policies change by term.

To check your own draft before you submit, there's our free AI detector. Because several of these policies lean on process evidence, it's also worth knowing what Google Docs replay tools actually record.

If your work is flagged

The policies above already build in the steps: notice, a conversation, more evidence than a score. Ask which of those your school uses, then ask for them. UT Austin's provost page, for one, says instructors "may not adjust or otherwise impact the student's grade on the academic assessment in question while the student is engaged in the conduct process."

You won't be the first. There's a record of students wrongly flagged by AI detectors, and a piece on what a single detector score can and cannot prove. The separate question of how accurate Pangram is has its own post.

FAQ

Which universities use Pangram?

As of October 27, 2026, UChicago makes it available to all instructors, and Dartmouth's Faculty of Arts and Sciences has authorized it for faculty use. Wellesley piloted it in fall 2025 and, per a March 2026 post by its CTO, opened it to any faculty member who wants it. Portland Community College is evaluating it. Check each school's own page for current status.

Does UT Austin use Pangram?

UT Austin's provost page says the university has no central contracts for AI detection software and doesn't allow third-party detectors on student work without a university contract or purchase order. The page doesn't mention Pangram.

Does Dartmouth use Pangram?

Dartmouth's Arts and Sciences faculty guidelines, adopted September 2026, name Pangram as the only authorized detector. The college hasn't decided whether to buy an institutional subscription, and faculty aren't required to use it.

Does my professor have to tell me they use Pangram?

That depends on your school. Dartmouth requires written notice, and UChicago's guidance tells instructors to state it clearly. Check your syllabus and your school's policy page.

If my school uses Canvas, does it have Pangram?

Not necessarily. Pangram lists Canvas and other platforms as integrations, but a school still has to license it and turn it on. At UChicago it works on individual assignments only.

How many universities use AI detectors?

In HumanizerNerd's check of 300 large schools, read October 9, 2026, 53 provided a detector, most often Turnitin, and only one used Pangram. Treat that as one site's sample, not a full count.