May 3, 2024
Top 20 US Schools Recommend Against AI Detectors
See Recommendation from Top 20 Schools in the US on AI Detectors
As the use of generative AI in education becomes more prevalent, many universities are grappling with how to address the issue of detecting AI-generated content in student work. The following table summarizes the recommendations and perspectives of several top U.S. universities regarding the use of AI detection tools:
| School | Recommendation | School Comment | Link |
|---|---|---|---|
| Princeton | Advise Against |
“We believe that detection and surveillance tools are not an effective means to identify or deter the use of generative AI. They’re not reliable, and they appear to be biased. We don’t recommend that faculty use these tools. ” | Read More |
| MIT | Advise Against |
“AI Detectors Don’t Work.” | Read More |
| Harvard | Advise Against |
“The FAS discouraged professors from using AI detection tools, which Stubbs said were too unreliable for use.” | Read More |
| Stanford | Don’t Recommend |
“Plagiarism detection platforms have varying levels of efficacy in detecting text generated by AI tools. Relying solely on plagiarism detection tools to identify if students have not complied with an AI-related course policy is not advised.” | Read More |
| Yale | Don’t Recommend |
“Controlling the use of AI writing through surveillance or detection technology is probably not feasible.” | Read More |
| UPenn | Strongly Advise Against |
“Avoid AI detectors. Several companies have created “AI detectors” that claim to evaluate whether text was written by an AI or a human. However, none of these tools are sufficiently accurate to serve as evidence that a student has used AI. Additionally, AI detectors demonstrate bias by falsely flagging written work by non-native English speakers. Uploading student work to an AI detector may also violate Penn’s AI privacy policies.” | Read More |
| Duke | Don’t Recommend |
“Learning Innovation does not endorse any software or programs that claim to determine if a student's writing was produced by AI.” | Read More |
| John Hopkins | Don’t Recommend |
“Advancements in generative AI technologies are outpacing development of tools designed to detect their usage. While numerous AI detection tools are available, testing indicates a wide range in the accuracy and efficacy of these tools—including missed AI content detection and false positives.” | Read More |
| Northwestern | Don’t Recommend |
“Turnitin piloted an AI Writing Indicator from April to August 2023, but after a series of consultations, Northwestern has decided to turn it off. We do not recommend using this detection tool as the basis for reporting a suspected case of academic dishonesty.” | Read More |
| Columbia | Caution Usage |
“As with any form of detection software, there are risks of misidentification, which can have consequences in the classroom.” | Read More |
| Cornell | Advise Against |
“We currently do not recommend using current automatic detection algorithms for academic integrity violations using generative AI, given their unreliability and current inability to provide definitive evidence of violations.” | Read More |
| University of Chicago | Don’t Recommend |
“While there are several tools purporting to be able to detect whether text has been AI-generated, studies of their efficacy have shown that they are unreliable, producing a large number of both false positives and negatives.” | Read More |
| UC Berkeley | Don’t Recommend |
“The AI detection provides no such conclusive proof, and it is unclear how these findings would or should be used in an academic integrity investigation.” | Read More |
| Rice | Don’t Recommend |
“The Honor Council is not currently planning to use AI detection software.” | Read More |
| Vanderbilt | Strongly Advise Against |
“We do not believe that AI detection software is an effective tool.” | Read More |
| University of Notre Dame | Don’t Recommend |
“Be aware of the limitations of currently available AI detectors. The detectors found online are not currently reliable enough to use as the basis for an accusation of academic dishonesty.” | Read More |
| University of Michigan | Caution Usage |
“AI detection tools should not be considered a definitive measure for cheating.” | Read More |
| Georgetown | Don't Recommend |
“Georgetown has turned off Turnitin’s AI writing detection feature because of the main campus faculty executive committee and the Honor Council’s concerns about the tool’s accuracy and in the belief that the harms of false positives are worse than any advantages of the tool.” | Read More |
| Carnegie Mellon University | Don't Recommend |
“Although companies such as Turnitin are beginning to offer AI detection services, none have been established as accurate.” | Read More |
| UNC | Caution Usage |
“Their accuracy is not guaranteed and there may be instances where they fail to detect plagiarism” | Read More |
Discover how Rumi supports AI literacy and academic integrity
Sign up for email updates
Keep up with Rumi news and product features.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Rumi® Technologies, Inc.
