Guidance on technology tools for academic integrity
Academic integrity is an important concern among instructors, particularly in light of new developments in generative AI technology. Some instructors may look to technological tools to address these concerns. This page discusses the many factors and issues related to academic integrity and technology tools at Stanford.
Current policy guidance related to technology
Integrity, honesty, and ethical behavior are fundamental to campus life at Stanford. With regard to teaching and learning, the Honor Code, Student Judicial Charter, and the guidance provided by the Board on Judicial Affairs available on the Office of Community Standards (OCS) website provides guidance and policy on matters involving cheating, what constitutes plagiarism, unpermitted aid, and so on.
Generative AI policy guidance
The Board of Judicial Affairs (BJA) has stated that individual instructors are free to set their own course policies regarding the use of generative AI tools.
OCS has stated: "Absent a clear statement from a course instructor, use of or consultation with generative AI shall be treated analogously to assistance from another person."
Therefore, instructors should decide their own course policies regarding the use of generative AI, state those policies in syllabi, and clearly communicate them to students.
OCS provides more details on its Generative AI Policy Guidance webpage.
Technology tools to enforce academic integrity
While there are many types of educational technology tools that might relate to academic integrity, most instructor inquiries are about plagiarism detection and proctoring.
Plagiarism detection tools
Tools, such as Turnitin, Unicheck, and Grammarly typically cross-reference uploaded student work to a database of other works to determine the originality and sources of the student work. Some plagiarism detection tools also leverage AI technology and may detect AI-generated text to varying degrees of accuracy.
Current policy guidance on plagiarism detection tools
Refer to OCS’s Tips for Faculty & Teaching Assistants web page for policy guidance on the use of plagiarism detection software.
How plagiarism detection software works
Students typically upload their work to a plagiarism detection platform in their campus learning management system. The platform checks the submitted work against a database of copyrighted or other external works and creates a report for each uploaded work that highlights any sections of writing or code that are similar to existing content. The report may also provide a link to the existing original content. This report is visible to the instructor and may be made visible to students. Uploaded works that are highly similar to other works are flagged for review.
How to respond to suspected plagiarism
If there is a concern that academic dishonesty may have occurred, for example when plagiarism detection tools flag a student’s work for potential plagiarism or improper citations, it is the instructor’s responsibility under the Honor Code to contact the Office of Community Standards. OCS staff will consult with you to determine the best way forward.
The instructor should not discuss the matter with the student ahead of consulting with OCS. Instructors may not penalize grades (reducing the student’s grade for engaging in suspected academic dishonesty before the student has been found responsible). Consulting with OCS staff does not obligate the person to file a concern; investigations do not begin unless and until an official written concern is filed.
This guidance applies to any assignment or activity that is used as the basis of grading. Suspected plagiarism on an ungraded assignment, such as a draft of an essay, may be resolved with the student directly and OCS does not need to be consulted.
In the case of suspected unpermitted aid from generative AI tools, verification through plagiarism detection tools is not required before submitting a concern to OCS.
After reporting a concern, the instructor will still have the opportunity to provide input on how the case is resolved. For example, the instructor could ask for a non-disciplinary resolution if the student has taken responsibility and the instructor feels the violation is minor.
Issues with plagiarism detection tools
Some plagiarism detection tools store uploaded student work which allows future submissions to be checked against the database. While some tools allow users to opt out or to keep separate institutional databases of submitted work, questions about maintaining the privacy and intellectual property rights of students should be addressed when adopting such tools.
For some students, submitting their work to plagiarism detection tools may erode feelings of trust and belonging. Some students may have had experiences in previous educational settings where such tools were associated with punitive measures, may have misconceptions about what plagiarism is, or not understand how plagiarism detection tools work. If unaddressed, this can contribute to distrust and disengagement in the classroom.
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.
Licensing costs for plagiarism detection tools can vary. Stanford currently does not support any campus-wide plagiarism detection tools. Individual schools, departments, or units within the university would cover any costs and provide instructor support if they choose to adopt such tools.
Benefits of plagiarism detection tools
These tools can be helpful when used thoughtfully to teach what plagiarism and academic integrity are and to practice writing skills such as researching, note-taking, paraphrasing, citing sources, and so on. Students might submit ungraded drafts of their writing to a plagiarism detection platform to identify areas to improve.
These tools can also be useful in non-instructional contexts, such as submitting works to a journal for publication.
Plagiarism detection tools are often bundled with other useful grading and writing tools, such as dynamic rubrics and grading annotation.
Remote proctoring tools
Online proctoring tools, such as Proctorio or ProctorU, typically use software installed on the students' computers to verify students’ identity, restrict certain apps, and activate the students’ web cameras to record or observe them while taking an exam. There are generally two types of online proctoring tools: ones that use automated software and ones that use live human proctors. Both will then report suspicious behavior to the instructor for follow-up.
Current policy guidance on proctoring tools
Policy around proctoring is evolving. The Office of Community Standards provides up-to-date information in the Honor Code section of its website.
Consider adopting these tools carefully
Stanford currently does not support any campus-wide plagiarism detection or proctoring tools for general use for many reasons.
Adoption of any academic technology tool at the campus-wide level, particularly tools that intersect with established policy, requires a careful and comprehensive process of evaluating the tool for security, privacy, accessibility, cost, technical support services, policy considerations, and so forth. Campus leaders are aware of these tools and the interest from some instructors in adopting them, and are continuously evaluating their feasibility. However, currently, such tools have not been adopted for general use.
Therefore, if you are in a specific situation with a compelling reason to use such tools, first contact the Office of Community Standards for the most up-to-date policy guidance. Then contact your department leaders to explore the adoption of these tools for specific use within your course, department, or program. Some departments have already adopted tools, notably the Computer Science department, for checking the originality of student work in programming languages.
Strategies to promote academic honesty
Regardless of the technology tools, it is important to remember that plagiarism detection and proctoring are based on a model of enforcement and punitive consequences that only address certain factors related to academic integrity.
A more inclusive and comprehensive approach might address factors that motivate ethical or unethical behavior, the deeper goals of education, and the human needs of instructors and students. For example, you might encourage students to consider the writing process as essential to their learning and work with them to foster the habits of academic integrity, such as accurately noting and acknowledging research sources.
Rather than only asking “How can we stop students from cheating?”, we might also ask, “How can we encourage students to behave honorably and ethically?”
For further discussion on this topic see Teaching Strategies to Support the Honor Code and Student Learning.