Teaching & Learning
- SDSU uses an assignment-level framework rather than one blanket rule for student coursework
- AI use in assessments and deliverables may be prohibited, restricted by instructor conditions, allowed with required documentation, or fully unrestricted depending on the classification assigned to the work
Classification Levels of Allowed Use of GenAI in Assessments and Deliverables: The following classification levels determine the permitted use of GenAI in-course assessments and deliverables:
● AI-1: Disallowed
● AI-2: Restricted
● AI-3: Documented
● AI-4: Unrestricted
Students submitting assessments and deliverables must adhere to specific transparency, documentation, and accuracy requirements. The classification of each assessment or deliverable will determine the level of AI use allowed:
AI-1 - Disallowed: Generative AI tools are strictly prohibited in this assessment or deliverable.
Any use of AI will be considered a violation of academic integrity.
AI-2 - Restricted: In this assessment or deliverable, students may use AI, but with restrictions and documentation requirements specified by the instructor.
AI-3 - Documented: Students can use AI in any manner for this assessment or deliverable, but they must provide appropriate documentation for all AI use.
AI-4 - Unrestricted: Generative AI tools can be used without restrictions, and documentation is not required.
- SDSU states that exam and assessment use of AI depends on the assigned AI classification for that assessment
- Separately, standard cheating rules also prohibit using unauthorized devices or materials during testing conditions
- If AI is disallowed, any use is an academic integrity violation; where AI is restricted or documented, students must follow instructor requirements
Classification Levels of Allowed Use of GenAI in Assessments and Deliverables: The following classification levels determine the permitted use of GenAI in-course assessments and deliverables:
● AI-1: Disallowed
● AI-2: Restricted
● AI-3: Documented
● AI-4: Unrestricted
Students submitting assessments and deliverables must adhere to specific transparency, documentation, and accuracy requirements. The classification of each assessment or deliverable will determine the level of AI use allowed:
AI-1 - Disallowed: Generative AI tools are strictly prohibited in this assessment or deliverable.
Any use of AI will be considered a violation of academic integrity.
AI-2 - Restricted: In this assessment or deliverable, students may use AI, but with restrictions and documentation requirements specified by the instructor.
AI-3 - Documented: Students can use AI in any manner for this assessment or deliverable, but they must provide appropriate documentation for all AI use.
AI-4 - Unrestricted: Generative AI tools can be used without restrictions, and documentation is not required.
using notes, cheat sheets, or other devices considered inappropriate under the prescribed testing condition;
- SDSU permits students to use GenAI for self-study and learning support
- The guidance specifically gives background research, self-study, and study scheduling as acceptable support uses, while also requiring users to validate accuracy and exercise judgment
Time management & organization: GenAI tools can help create study schedules and help organize notes.
Example: I acknowledge using [1] ChatGPT to [2] generate materials for background research and self-study to create this assignment.
Validate the accuracy of information sourced from GenAI by cross-referencing it with trusted resources.
Remember that GenAI tools are not infallible and should not be the sole basis for significant decisions.
Research
- In human subjects research, investigators may use AI to draft consent documents, but the PI must review them for accuracy and clarity
- For graduate theses and dissertations, SDSU does not set one uniform campus rule; Graduate Council strongly recommends each master's and doctoral program develop its own AI policy
- Across those graduate research contexts, theses and dissertations must remain the student's original contribution, AI use must be transparent, and students remain responsible for proper acknowledgment and verification
The San Diego State University Graduate Council strongly suggests that each Master’s and Doctoral program develop guidelines and/or policies regarding student use of Artificial Intelligence and related data processing (hereafter “AI”), based on the general principles listed below.
It is recommended that the resulting document explain in a student-centric manner the appropriate and inappropriate uses of AI in Master’s theses, Doctoral dissertations, other major program milestones, research, publications, coursework, and/or other professional spaces.
Theses and dissertations must reflect a student’s original contribution. Students must clearly and transparently distinguish their own work from that of prior scholars. The use of generative AI tools does not absolve the student from this responsibility. Permissible uses of AI vary among degree programs, as well as the appropriate citation model.
Using AI does not exempt students from plagiarism or academic honesty standards. Any use of AI tools must be transparent. Students are fully responsible for maintaining integrity and properly acknowledging the original source of an intellectual contribution.
If using AI to create consent documents, the PI must review the document for accuracy and clarity, as all required elements may not be included in the resulting consent document, or the consent document may not be consistent with the protocol.
- In human subjects research, SDSU requires researchers to use only SDSU-licensed AI platforms and prohibits uploading PL-1 or PL-2 data into AI platforms
- If AI or machine learning is used, the IRB application must explain how outputs will be reviewed or validated by the PI or research staff, and updated iRIS requirements ask researchers to specify purpose, data types, tools, bias mitigation, and human oversight
Use only SDSU-licensed AI platforms (e.g., CoPilot, ChatGPT, Gemini).
Do not upload PL-1 (i.e., Protected Health Information, name with personally identifiable information, SSN, law enforcement records) or PL-2 (i.e., information covered under FERPA, such as grades, courses taken, test scores) data into an AI platform.
As AI data training contains human bias, the “Use of Artificial Intelligence (AI) or Machine Learning” section of the IRB application must include a description of how AI results/outputs will be reviewed or validated by the PI or research staff.
A new guidance document on the Use of Artificial Intelligence (AI) in Human Subjects Research is now available. This guidance outlines important requirements and considerations for researchers who use AI tools at any stage of study design, data analysis, or document preparation.
Requirements include:
* Purpose of Use: Explain why AI is being used and the specific tasks it will perform (e.g., data analysis, recruitment, or drafting materials).
* Data Specifics: Specify the type of data involved (identifiable, coded, or de-identified).
Tools & Platforms: List all specific tools, algorithms, and platforms (e.g., SDSU-licensed versions of ChatGPT, CoPilot, or Gemini).
* Bias & Fairness: Describe steps taken to minimize algorithmic bias and ensure equitable results.
* Human Oversight: Detail how the PI or research staff will review and validate AI-generated outputs.
- For human subjects research, SDSU requires explicit IRB and consent-related disclosure and oversight when AI is used
- Graduate research guidance also states that AI use does not exempt students from plagiarism or academic honesty standards and must be transparent
- Researchers must describe in the IRB application how AI outputs will be reviewed or validated, and participants must be informed in the consent document about how AI will be used in the study
As AI data training contains human bias, the “Use of Artificial Intelligence (AI) or Machine Learning” section of the IRB application must include a description of how AI results/outputs will be reviewed or validated by the PI or research staff.
In the consent document, researchers must inform participants about how AI will be used in the study.
Using AI does not exempt students from plagiarism or academic honesty standards. Any use of AI tools must be transparent. Students are fully responsible for maintaining integrity and properly acknowledging the original source of an intellectual contribution.
Academic Integrity
- Graduate guidance also requires AI use to be transparent and properly acknowledged
- The summary statement is expected to identify the AI system and link, the purpose of use, the prompts used, and how the output was used or modified
- SDSU requires disclosure of AI use when an assessment is classified AI-2 or AI-3, with documentation either through primary source citations or a summary statement if the instructor allows that form
Standards for acknowledging the use of GenAI when allowed: When assessments or deliverables fall under AI-2 or AI-3 classifications, students must provide documentation regarding their use of generative AI. This documentation can take the form of primary source citations or summary statements, as specified by instructors.
In cases where primary source citations are mandated, all factual statements in the work must be attributed to the original sources following standard prescribed citation styles.
Summary Statement: For specific assignments, providing a summary statement that discloses the utilization of generative AI and describes its application may suffice. These summary statements should include details such as the name and URL of the AI system, the specific purpose for which generative artificial intelligence was employed, the prompts that were used, and how the AI-generated output was utilized or modified:
I hereby acknowledge the use of [insert AI system(s) and link] for [specify the particular use of generative artificial intelligence]. The prompts employed in this process encompass [list of prompts]. The outcomes generated by these prompts were utilized to [clarify the use].
Any use of AI tools must be transparent. Students are fully responsible for maintaining integrity and properly acknowledging the original source of an intellectual contribution.
- SDSU explicitly treats unauthorized AI use as academic misconduct in two ways
- The provided sources do not state a university position on AI detection tools
- For work classified AI-1, any AI use is an academic integrity violation, and the student conduct page separately defines representing work produced by generative AI as one's own as cheating/plagiarism
AI-1 - Disallowed: Generative AI tools are strictly prohibited in this assessment or deliverable.
Any use of AI will be considered a violation of academic integrity.
representing work produced by generative artificial intelligence (AI) as one's own.
Institutional & Administrative
- Faculty and staff are encouraged to complete the AAAI Micro-Credential if they intend to use GenAI tools
- Initial drafts should never be treated as final products, and facts and data must be reviewed and validated before sharing
- SDSU permits faculty and staff to use GenAI to support work such as brainstorming, drafting, revising, and communications, but it must augment rather than replace original work and human expertise
SDSU students, faculty and staff interested in utilizing GenAI are encouraged to adopt GenAI tools appropriately and ethically. Such tools should augment or support original work and not be autonomous, and SDSU professionals should strictly follow the guidelines below.
GenAI tools can support SDSU students, faculty and staff in many ways during several phases of their work:
Brainstorming ideas and planning: Before creating content, GenAI may be used to brainstorm topics, questions, ideas and concepts, and also to collect examples, develop outlines, identify themes or organize plans.
Drafting content: Some GenAI tools support generating and editing content based on prompts, including initial and draft versions of scripts, letters, translations into multiple languages, descriptions, and social media posts.
Revising: After original content is generated, many GenAI tools can aid in enhancing clarity, structure and grammar in writing (while ensuring original thought and citation).
Communications & Reports: To complement your work, GenAI can assist in drafting internal memos or refining emails.
Regardless of the tool adopted or phase of work, initial drafts should never be presented as a final product in the creation of class assignments, proposals, emails, articles, research presentations, brochures, flyers, fact sheets, media releases, or other information, whether for an internal or external audience. Data, facts, and other information must continue to be rigorously reviewed and validated before being published or shared.
Faculty and staff who intend to use GenAI tools are encouraged to complete the Academic Applications of AI (AAAI) Micro-Credential.
- The university also warns that most commercial AI systems do not comply with HIPAA or FERPA protections
- SDSU directs its community to use professional or SDSU-licensed AI tools and provides ChatGPT Edu for all employees and students with enterprise privacy and security protections
- Public AI tools must not receive confidential, proprietary, embargoed, PL-1, or PL-2 data, and uploaded material should be non-sensitive, de-identified where applicable, and shareable with permission
It's important to note that most commercially available AI systems do not comply with HIPAA or FERPA protections.
Further, care should be taken to use professional versions of GenAI tools. At SDSU, ChatGPT Edu, Google Gemini and
Never upload confidential, proprietary or embargoed information or data to public GenAI tools (e.g., data from Canvas, my.SDSU and other examples provided in this document's Human Oversight and Review section).
OpenAI's ChatGPT Edu Agreement is tailored for the California State University (CSU) system, providing advanced AI capabilities for all employees and students. It offers the ability to build custom GPTs for sharing within campus workspaces, as well as the needed privacy and data protection, and other enterprise security such as single sign-on (SSO).
All SDSU students, faculty, and staff can activate and access ChatGPT Edu using their SDSUid username and password (no separate signup or password required).
Our agreement with OpenAI includes language that ensures your ChatGPT interactions and data are not used to train their underlying large language models or improve their services.
Yes, the CSU ChatGPT Edu plan includes enterprise commercial data protection features like SAML Single Sign-On (SSO), domain verification, custom data retention windows, SOC 2 Type 2 compliance, and data encryption at rest and in transit.
You can upload:
* Non-sensitive data that does not contain Level 1 (L1) classified information.
* Public or openly shareable data that does not require special access controls.
* De-identified data that has been stripped of personally identifiable information (PII) or other sensitive details.
* Non-copyrighted material that you have permission to share and process.
Use only SDSU-licensed AI platforms (e.g., CoPilot, ChatGPT, Gemini).
Do not upload PL-1 (i.e., Protected Health Information, name with personally identifiable information, SSN, law enforcement records) or PL-2 (i.e., information covered under FERPA, such as grades, courses taken, test scores) data into an AI platform.
- SDSU has institution-level AI guidance and training infrastructure rather than a single binding governance code in the provided sources
- In graduate education, Graduate Council has approved general principles and strongly recommends that each master's and doctoral program create its own AI guidelines or policies
- The GenAI guidance frames responsible, ethical, and secure adoption across students, faculty, and staff, and the Academic Applications of AI Micro-Credential is identified as a university training resource
SDSU students, faculty and staff should follow important considerations in adopting any GenAI tools related to information security, data and personal privacy, compliance, copyright and protecting proprietary information, accuracy, and
Training and upskilling opportunities are essential for all students, faculty and
At SDSU, the Academic Applications of AI (AAAI) Micro-Credential is a freely accessible training.
Graduate Council has approved a set of general principles to guide policy in these areas, and strongly recommends that each Master’s and Doctoral program develop their own guidelines and/or policies.
The San Diego State University Graduate Council strongly suggests that each Master’s and Doctoral program develop guidelines and/or policies regarding student use of Artificial Intelligence and related data processing (hereafter “AI”), based on the general principles listed below.
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 San Diego State University (SDSU)'s AI Policies
San Diego State University (SDSU) has defined AI policies in 11 of 12 categories, with an overall coverage score of 92%.
SDSU requires disclosure of AI use when an assessment is classified AI-2 or AI-3, with documentation either through primary source citations or a summary statement if the instructor allows that form. The summary statement is expected to identify the AI system and link, the purpose of use, the prompts used, and how the output was used or modified. Graduate guidance also requires AI use to be transparent and properly acknowledged.
SDSU explicitly treats unauthorized AI use as academic misconduct in two ways. For work classified AI-1, any AI use is an academic integrity violation, and the student conduct page separately defines representing work produced by generative AI as one's own as cheating/plagiarism. The provided sources do not state a university position on AI detection tools.
SDSU directs its community to use professional or SDSU-licensed AI tools and provides ChatGPT Edu for all employees and students with enterprise privacy and security protections. Public AI tools must not receive confidential, proprietary, embargoed, PL-1, or PL-2 data, and uploaded material should be non-sensitive, de-identified where applicable, and shareable with permission. The university also warns that most commercial AI systems do not comply with HIPAA or FERPA protections.