Teaching & Learning
- AI use for coursework and assignments is determined by the individual course instructor
- Instructor course policies supersede general university guidance; if no course policy is stated, students should ask the instructor before using AI tools
“Your course instructor’s policy on the use of AI tools always supersedes any general university guidance. In the absence of a stated course policy about AI, students should ask their instructors about their expectations for if or how to use it.”
- Use of generative AI in examinations and other assessments is governed by the course instructor’s policy
- Using AI when it is not explicitly permitted by the instructor can be treated as a violation of Tufts’ Academic Integrity policy
“The use of AI when not explicitly permitted by a course instructor can be considered a violation of Tufts’ Academic Integrity policy.”
- Tufts provides permissive guidance for students to use generative AI for learning and study assistance (e.g., summarizing text, generating study guides/practice questions, explaining concepts, and brainstorming), subject to any course-specific restrictions
“Students can use AI to ... summarize complex text, generate study guides or questions, and explain concepts in different ways. Many generative AI tools can serve as a brainstorming partner... [and] offer a good starting point to generate a wide range of ideas.”
- AI-generated code is subject to course/instructor rules and Tufts academic integrity expectations
- Submitting code authored by another source without acknowledgement (which can include AI-generated programs) may constitute plagiarism; work submitted to satisfy an academic requirement must be the student’s own
“Plagiarism is the unacknowledged use of the work of another... It also includes the submission of a computer program that has been authored by another... Work submitted to satisfy an academic requirement must be the student’s own.”
Research
- Researchers may use AI tools for tasks like summarizing documents and generating ideas, but must follow relevant publisher/journal rules and remain responsible for the accuracy of any AI-generated text and the integrity of their scholarship
“Some tools... can analyze uploaded documents and help a researcher summarize them or generate ideas for new scholarship... Many publishers prohibit the use of AI-generated content in part or in full, and some journals specify how it can be used ethically. Ultimately, you are responsible for checking the accuracy of any generated text and for the integrity of your scholarship.”
- Confidential, protected, or proprietary information (including data classified at Level 2 or above under Tufts’ information classification policy) must not be entered into public generative AI tools, including for research data analysis
“Do not enter confidential, protected, or proprietary information into public generative AI tools... This can include student records, employee information, patient data, intellectual property, or any other data classified at Level 2 or above according to Tufts’ Information Classification and Handling Policy.”
- Research misconduct is defined to include fabrication, falsification, or plagiarism in proposing, performing, reviewing, or reporting research
- Use of generative AI in research is subject to Tufts research integrity expectations; researchers remain responsible for verifying AI outputs and for the integrity of their scholarship
“Ultimately, you are responsible for checking the accuracy of any generated text and for the integrity of your scholarship... Research Misconduct is defined as fabrication, falsification, or plagiarism in proposing, performing, or reviewing research, or in reporting research results.”
Academic Integrity
- When AI use is permitted in a course, students are responsible for properly citing how they used the tool, consistent with instructor expectations; if no course policy is stated, students should ask the instructor about expectations for AI use and documentation
“In the absence of a stated course policy about AI, students should ask their instructors about their expectations for if or how to use it. Students are responsible for properly citing how they have used the tool.”
- Unauthorized AI use (when not explicitly permitted by the course instructor) can be treated as an academic integrity violation
- Tufts advises against using AI text-detection tools due to questions about reliability and documented bias, and notes that some detection sites are blocked on the Tufts network
“The use of AI when not explicitly permitted by a course instructor can be considered a violation of Tufts’ Academic Integrity policy.” and “Because of significant questions and documented bias in AI text-detection tools, use of these sites is not recommended at this time (and some are blocked on the Tufts network).”
Institutional & Administrative
- Faculty and staff using AI tools must protect university data and should not enter confidential/protected/proprietary information (including Level 2+ data) into public generative AI tools
“Anyone using these tools has a responsibility to protect university data. This can include student records, employee information, patient data, intellectual property, or any other data classified at Level 2 or above according to Tufts’ Information Classification and Handling Policy.”
- Tufts prohibits entering confidential/protected/proprietary information (including Level 2+ data) into public generative AI tools
- Tufts provides access to Microsoft Copilot with commercial data protection, and blocks unapproved third-party AI bots/apps from being added to Tufts-licensed Zoom and MS Teams
“Do not enter confidential, protected, or proprietary information into public generative AI tools.” and “Microsoft Copilot is available to the Tufts community and provides commercial data protection.” and “Tufts has blocked unapproved 3rd party AI related bots/apps from being added to Tufts licensed Zoom and MS Teams.”
- Tufts’ AI guidelines were developed by an informal cross-unit AI Working Group (with CELT and Tufts Technology Services) and are intended to be updated frequently as the university continues to gather community experience and feedback on AI use
“The following guidelines were developed by an informal AI Working Group, in collaboration with colleagues in the Center for Enhancement of Learning and Teaching (CELT) and in Tufts Technology Services (TTS).” and “The guidelines are evolving and we expect to update them frequently as we continue to talk with faculty, students, and staff about their experiences with AI.”
DocuMark: Responsible AI Use for Academic Integrity
Knowing your institution's AI policy is step one. DocuMark helps enforce it fairly by empowering universities to manage AI-generated content, prevent cheating, and support student writing through responsible AI use.
Common Questions About Tufts University's AI Policies
Tufts University has defined AI policies in 12 of 12 categories, with an overall coverage score of 100%.
When AI use is permitted in a course, students are responsible for properly citing how they used the tool, consistent with instructor expectations; if no course policy is stated, students should ask the instructor about expectations for AI use and documentation.
Unauthorized AI use (when not explicitly permitted by the course instructor) can be treated as an academic integrity violation. Tufts advises against using AI text-detection tools due to questions about reliability and documented bias, and notes that some detection sites are blocked on the Tufts network.
Tufts prohibits entering confidential/protected/proprietary information (including Level 2+ data) into public generative AI tools. Tufts provides access to Microsoft Copilot with commercial data protection, and blocks unapproved third-party AI bots/apps from being added to Tufts-licensed Zoom and MS Teams.