Teaching & Learning
- MIT also provides an example syllabus statement recommending that, to avoid plagiarism issues, instructors ask students to cite AI-generated material and reflect on how they used and modified GenAI output
- MIT does not define a single Institute-wide student policy for whether generative AI may be used to complete graded coursework; instead, guidance emphasizes that instructors should clearly communicate subject-level expectations and consequences in the syllabus
Regardless of your thoughts on using GenAI in your subject, convey those thoughts, the resultant subject policies, and the consequences of their violation with your students at the beginning of the semester.
Including an AI Acceptable-use Statement in your syllabus and discussing your rationale for the policy on the first day of class can help regulate students' use of GenAI and open the door for additional conversations about GenAI and learning in your subject as the semester progresses.
However, to help you learn to use GenAI in ways that support your growth as a learner and teacher, and to avoid any issues of plagiarism – I ask that you: (a) include a citation for any material or idea that was AI generated; (b) include in your assignment a reflection of how you used GenAI, and how you modified its output to complete your assignment;
- MIT provides instructor-focused resources about redesigning assignments and assessments in light of generative AI, but does not state a universal allowance or prohibition for AI use during exams
- MIT does not define an Institute-wide rule for student AI use during examinations or assessments; guidance focuses on instructors setting and communicating subject-specific AI policies and consequences
Regardless of your thoughts on using GenAI in your subject, convey those thoughts, the resultant subject policies, and the consequences of their violation with your students at the beginning of the semester.
Including an AI Acceptable-use Statement in your syllabus and discussing your rationale for the policy on the first day of class can help regulate students' use of GenAI and open the door for additional conversations about GenAI and learning in your subject as the semester progresses.
- MIT does not define an Institute-wide policy specifically governing student use of generative AI for non-graded learning or study assistance
- Instructor guidance emphasizes that expectations about GenAI use should be conveyed via subject-level policies and an AI acceptable-use statement in the syllabus
Regardless of your thoughts on using GenAI in your subject, convey those thoughts, the resultant subject policies, and the consequences of their violation with your students at the beginning of the semester.
Including an AI Acceptable-use Statement in your syllabus and discussing your rationale for the policy on the first day of class can help regulate students' use of GenAI and open the door for additional conversations about GenAI and learning in your subject as the semester progresses.
- IS&T notes that certain MIT-licensed generative AI developer tools are available to MIT faculty and staff, which may be used for development purposes subject to IS&T guidance
- MIT does not define an Institute-wide student policy on use of AI tools for code generation in programming coursework; guidance presented focuses on instructors communicating subject-level expectations, policies, and consequences
Regardless of your thoughts on using GenAI in your subject, convey those thoughts, the resultant subject policies, and the consequences of their violation with your students at the beginning of the semester.
The products listed here provide or include generative artificial intelligence (AI) tools, services, or components, and are licensed by IS&T for use by the MIT community.
Azure OpenAI Developer access to Azure OpenAI and other cognitive services (e.g., Whisper, DALL-E) MIT faculty and staff Institute cost object required Register for an Azure account through IS&T's MIT Cloud Accounts program
Research
- MIT's IS&T guidance on generative AI emphasizes protecting non-public research outputs and unpublished research papers from being entered into generative AI tools
- This guidance frames generative AI use decisions around intellectual property, confidentiality, and academic integrity, and explicitly lists unpublished research papers and non-public research results/data as inappropriate for generative AI tools
Information Systems and Technology (IS&T) is providing this initial guidance to encourage community members to consider factors including information security, data privacy, regulatory and policy compliance, compliance with confidentiality restrictions concerning third party information and data, intellectual property (e.g., copyright and patent), and academic integrity when choosing to use or purchase software that makes use of generative artificial intelligence (AI).
Examples of information inappropriate for generative AI tools include non-public research results and data, unpublished research papers, confidential information received from third parties (such as research sponsors and collaborators), unpublished invention disclosures and patent applications, Institute financial and human resources information, personally identifiable information (including, for example, student records, medical records), any information subject to legal or regulatory requirements necessitating its proper safeguarding and handling, and any other information not intended to be freely available to the general public, or to the MIT community without access controls.
- COUHES guidance for research involving AI highlights risks to data privacy/confidentiality and includes consent-form considerations (including data sharing with an AI model and limits on withdrawal once shared)
- MIT's IS&T guidance states that non-public research results and data are inappropriate for generative AI tools, and it highlights confidentiality, privacy, and regulatory compliance considerations when using generative AI
Examples of information inappropriate for generative AI tools include non-public research results and data, unpublished research papers, confidential information received from third parties (such as research sponsors and collaborators), unpublished invention disclosures and patent applications, Institute financial and human resources information, personally identifiable information (including, for example, student records, medical records), any information subject to legal or regulatory requirements necessitating its proper safeguarding and handling, and any other information not intended to be freely available to the general public, or to the MIT community without access controls.
Artificial Intelligence (AI) is an ever-evolving field that presents new opportunities for research while amplifying the risk of data privacy and confidentiality.
If research data will be used to further develop the AI Model and if subject's data will be used to develop other AI Models. The consent form must include the impact of data sharing broadly with an AI Model and that once shared, information cannot be withdrawn.
- COUHES also directs investigators to review MIT IS&T's guidance on generative AI to ensure they are following MIT best practices for generative AI use
- COUHES provides AI-in-research guidance emphasizing privacy and confidentiality risks and instructs investigators to consider how AI use affects informed consent, data sharing, data storage, and commercial AI software use
Artificial Intelligence (AI) is an ever-evolving field that presents new opportunities for research while amplifying the risk of data privacy and confidentiality.
When appropriate, consider the following information when submitting a new application involving AI:
If the research uses commercial AI software, include if subject data will be stored and used to further develop the product.
Investigators must also consider any additional policies, such as funding, laws, legal or journal requirements, that may apply to the research. In addition to the information provided above, investigators should review the guidance on AI provided by MIT's IS&T: Initial guidance for use of Generative AI tools. This guidance will help ensure investigators are following MIT's best practices for the use of Generative AI.
Academic Integrity
- MIT does not have a formal Institute-wide policy on AI citation
- However, instructor guidance provides a sample syllabus statement that can be used to ask students to cite AI-generated material and reflect on its use
- Additionally, MIT Libraries offers a guide on how to cite AI tools like ChatGPT, noting that citation practices are still emerging and providing specific format examples for students and researchers to follow
However, to help you learn to use GenAI in ways that support your growth as a learner and teacher, and to avoid any issues of plagiarism – I ask that you: (a) include a citation for any material or idea that was AI generated; (b) include in your assignment a reflection of how you used GenAI, and how you modified its output to complete your assignment.
How to cite these tools is still an emerging area. Until citation standards are established, we recommend that you both check with your instructor, and provide a clear and detailed account of how you used the tool... Example: OpenAI. (2023). ChatGPT (Mar 14 version) [Large language model]. https://chat.openai.com/chat
- MIT does not define AI detection or enforcement specifics (e.g., use of AI detectors) in the provided source text
- The Academic Integrity Handbook PDF states that the MIT Policy on Student Academic Dishonesty is outlined in MIT's Policies and Procedures 10.2, but the provided policy pages did not contain extractable enforcement text within the accessible content retrieved here
The MIT Policy on Student Academic Dishonesty is outlined in MIT's Policies and Procedures 10.2
at http://web.mit.edu/policies/10/10.2.html.
Institutional & Administrative
- MIT also provides faculty-facing teaching guidance emphasizing that instructors should communicate subject policies and consequences regarding GenAI use through the syllabus and early-semester discussions
- MIT's IS&T issues generative AI guidance to the MIT community focused on information security, privacy, confidentiality restrictions, intellectual property, and academic integrity when choosing to use or purchase generative AI software
Information Systems and Technology (IS&T) is providing this initial guidance to encourage community members to consider factors including information security, data privacy, regulatory and policy compliance, compliance with confidentiality restrictions concerning third party information and data, intellectual property (e.g., copyright and patent), and academic integrity when choosing to use or purchase software that makes use of generative artificial intelligence (AI).
Regardless of your thoughts on using GenAI in your subject, convey those thoughts, the resultant subject policies, and the consequences of their violation with your students at the beginning of the semester.
- MIT's IS&T warns against using publicly available tools not subject to an Institute licensing agreement for MIT research and educational activities, and provides a list of MIT-licensed generative AI tools available to the community
- IS&T also specifies examples of information inappropriate for generative AI tools (including non-public research data/results, unpublished research papers, third-party confidential information, unpublished invention disclosures/patent applications, Institute financial/HR information, and personally identifiable information such as student and medical records) and advises contacting ai-guidance@mit.edu for an assessment if a desired tool is not on the approved list
We do not recommend you download, install, or otherwise use publicly available tools not subject to an Institute licensing agreement for your MIT research and educational activities.
The products listed here provide or include generative artificial intelligence (AI) tools, services, or components, and are licensed by IS&T for use by the MIT community.
If a tool or service you wish to use is not included below, contact ai-guidance@mit.edu to request an assessment before moving forward with use or purchase.
Examples of information inappropriate for generative AI tools include non-public research results and data, unpublished research papers, confidential information received from third parties (such as research sponsors and collaborators), unpublished invention disclosures and patent applications, Institute financial and human resources information, personally identifiable information (including, for example, student records, medical records), any information subject to legal or regulatory requirements necessitating its proper safeguarding and handling, and any other information not intended to be freely available to the general public, or to the MIT community without access controls.
- The provided sources do not define an Institute-wide AI governance structure (e.g., committees) or an overarching institutional AI strategy/roadmap beyond this initial guidance and tooling direction
- MIT's IS&T states it has assembled initial guidance and a list of MIT-licensed generative AI tools and services to assist community members in using generative AI while protecting MIT information and adhering to Institute policies and best practices
With MIT community members expressing increased interest in using generative artificial intelligence (AI) and considering the Institute's breadth of expertise in this rapidly-evolving field, Information Systems and Technology (IS&T) has assembled this initial guidance and list of MIT-licensed tools and services to assist community members in making use of generative AI, while protecting MIT information and adhering to Institute policies and best practices.
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 Massachusetts Institute of Technology's AI Policies
Massachusetts Institute of Technology has defined AI policies in 12 of 12 categories, with an overall coverage score of 100%.
MIT does not have a formal Institute-wide policy on AI citation. However, instructor guidance provides a sample syllabus statement that can be used to ask students to cite AI-generated material and reflect on its use. Additionally, MIT Libraries offers a guide on how to cite AI tools like ChatGPT, noting that citation practices are still emerging and providing specific format examples for students and researchers to follow.
MIT does not define AI detection or enforcement specifics (e.g., use of AI detectors) in the provided source text. The Academic Integrity Handbook PDF states that the MIT Policy on Student Academic Dishonesty is outlined in MIT's Policies and Procedures 10.2, but the provided policy pages did not contain extractable enforcement text within the accessible content retrieved here.
MIT's IS&T warns against using publicly available tools not subject to an Institute licensing agreement for MIT research and educational activities, and provides a list of MIT-licensed generative AI tools available to the community. IS&T also specifies examples of information inappropriate for generative AI tools (including non-public research data/results, unpublished research papers, third-party confidential information, unpublished invention disclosures/patent applications, Institute financial/HR information, and personally identifiable information such as student and medical records) and advises contacting ai-guidance@mit.edu for an assessment if a desired tool is not on the approved list.