AI Guidelines

CSU East Bay recognizes that generative artificial intelligence (GenAI) is already embedded in higher education, the workforce, and everyday institutional operations. The purpose of this framework is to articulate CSUEB values-aligned guidance for ethical, responsible, and mission-consistent use of generative AI across campus.

Responsible use of generative AI at work means employing AI tools transparently and ethically to support efficiency, learning, and decision-making, while preserving human judgment, accountability, privacy, and institutional values.

This guidance is informed by extensive campus town-hall listening sessions held during Fall 2025 and reflects shared concerns and priorities expressed by CSUEB faculty, staff, students, and administrators regarding the rapid adoption of generative AI technologies.

Without locking into tool-specific commitments, this guidance establishes campus-wide expectations for the responsible use of generative artificial intelligence (GenAI). It aligns with CSU system principles on academic integrity, data protection, accessibility, and ethical use of technology, while supporting innovation and efficiency.

Core principles

• Generative AI may be used as a supportive tool, not as a substitute for human judgment, expertise, or responsibility. Individuals remain fully accountable for all AI-assisted work products, decisions, and outcomes.
• Safe use of AI means matching the tool to appropriate tasks, and understanding what generative AI can and cannot do. Users need to be aware of risks arising from overreliance, bias in AI training data, AI’s lack of contextual or emotional understanding, and potential tendency to fabricate plausible but false information.
• Ethical use of AI prioritizes transparency, data protection, academic freedom, and human responsibility, supported by ongoing education and shared governance.

Applicability Preview

Students may use generative AI only in ways permitted by course policies and academic integrity standards. AI may support learning but must not replace required intellectual work.

Employees (faculty and staff/administrators) may use generative AI to support teaching, research, and operations, provided such use complies with CSU and state policies on data security, privacy, accountability, and professional responsibility. AI must not be used to replace professional judgment, decision-making authority, or compliance obligations.

Guidelines

Section I. Ensuring Data Security, Confidentiality, and Safe Use

Generative AI tools can inadvertently capture, store, or reuse sensitive information. This guidance reinforces the need to follow CSU data classification standards, use approved systems when handling protected data, and avoid practices that could compromise student, employee, or institutional information. CSUEB follows CSU Information Security Responsible Use Policy and California laws relating to GenAI.

1. Mandatory Use of Licensed Tools and Data Use Prohibitions:
   1. All campus users must use only institutionally vetted and accessible GenAI tools (e.g., ChatGPT EDU, Microsoft Copilot, or equivalents approved through ITC review).
   2. CSUEB explicitly prohibits the use of public/consumer versions of AI tools for any institutional work, even non-sensitive tasks, to prevent accidental data leakage.
   3. Users are strictly prohibited from inputting Level 1 (highly sensitive) data, personally identifiable information (PII), or confidential institutional data (including FERPA, HIPAA, or grant-related data) into any GenAI tool, regardless of license.
   4. Users can stay informed by completing self-paced CSU AI-related data and information security training.
2. Psychological and Physical Safety Protocols: CSUEB is committed to creating and sustaining an educational and working environment free of misconduct, stalking, threat, violence or any method including digital co-dependency that may cause or create a risk of physical or psychological injury to any current, former or prospective campus community member. (APA Health Advisory on Use of GenAI Chatbots and Wellness Applications for Mental Health – Nov 2025) (PDF file Download - Sweety please provide) (policy summary-intro statement). How-to/implementation below?
   1. CSUEB shall explore technical safeguards, such as internal notification triggers or query restrictions, to identify and redirect users who enter prompts suggestive of self-harm or crisis, linking them directly to campus support resources.
   2. Ensuring psychological and physical safety of our campus communities remains a priority through integrating digital wellness and AI safety training into mandatory student orientation, first-year experience, and other ongoing training programs.
   3. Make ongoing resources available through counseling and student health services.
   4. Reviewing and expanding protocols to include regular, evidence-based assessments of the psychological and social impacts of AI adoption as the technology advances.
   5. Any academic program or course assignment involving use of AI to create bots or innovative products must include consideration of and statement on impact on personal and community safety and well-being.
      
      
3. Accessibility Compliance:
   
   1. All AI tools adopted or licensed by the institution must comply with WCAG 2.2 AA and Section 508 standards to ensure equitable access for all users.
   2. Purchasing or and requisition procedure for new ICT products shall include accessibility compliance.
      
      
4. Data Storage and Retention Policy:
   1. CSUEB thru unit, department and college leads shall explicitly define how AI interaction logs or generated content will be stored, who can access them, and how long they are retained, ensuring FERPA and GDPR alignment.
   2. Data storage and retention policy shall be reviewed and updated annually by ITS and in consultation with ITC for Academic Technology and ITS unit assigned individual(s).
      
      
5. Student Data Consent:
   1. Require explicit informed consent for any AI system that collects or processes identifiable student data, consistent with FERPA §99.30 and emerging state privacy laws.
   2. Establish holding place in system-procedure
      
      
6. AI Procurement and Vendor Review:
   1. All new AI tools proposed for institutional use must undergo a formal review process for data security, algorithmic bias, and accessibility compliance before procurement or integration into campus systems.
   2. The review process may benefit from target user participation in a limited trial before full launch. This step will benefit from being added to the current ITC review process.

Section II. Advancing Academic Integrity, Innovation, and Pedagogy

As we continue to explore the implications (ethical, societal, environmental, legal, and practical) of AI uses and advancements, CSUEB is committed to prepare all users especially students to critically use and critically assess outputs for authenticity and well-being of all members.

1. Preservation of Core Learning and Critical Thinking:
   
   1. GenAI tools must be treated as supplemental learning aids, similar to spellcheck or calculators, not replacements for fundamental intellectual effort and processes, such as critical thinking, brainstorming, or original writing.
   2. Users remain solely accountable for the accuracy, originality, and content of their final work.
2. Required Disclosure and Attribution:
   1. All academic submissions, research papers, grants, and administrative reports that rely substantially on GenAI must include a disclosure statement specifying which AI tool was used, for what purpose, and to what extent (e.g., brainstorming, editing, data analysis).
   2. For internal use and audiences, the disclosure format may be standardized institution-wide to ensure consistency.
      
      • Here are some suggestions from Cleland et al (2026).
      • When using GenAI in academic work, what is the difference between citation and acknowledgement? (Arizona State University)
   3. Disclosures must also include an accessibility statement if AI tools were used to generate alternate formats (e.g., captioning, alt text, transcripts).
   4. If AI use is allowed in an academic assignment, appropriate acknowledgement is required. Here are some guidelines from CSUEB Library for citing ChatGPT
      • MLA - How do I cite generative AI in MLA style (updated August 2025)
      • APA - Citing generative AI in APA Style (Updated September 2025)
      • Chicago Manual - How do you recommend citing content developed or generated by artificial intelligence, such as ChatGPT?
3. Ethical Assessment Practices:
   1. Faculty shall not use AI tools as the sole mechanism to grade student work or generate personalized feedback without meaningful human review and judgment. AI may be used to assist with grading logistics (e.g., pattern identification, preliminary scoring) but final assessment decisions must involve direct faculty evaluation.
   2. Faculty reverting to exclusively in-person, timed, or handwritten assessments to prevent AI use, must make alternate arrangements for students requiring accommodations.
4. Course and Syllabus Policy:
   25-26 CIC 1 (Aug 24, 2025) was approved by Senate and President Sandeen on Oct 2, 2025. This update the Syllabus Policy 22-23 CIC 6/22-23 FAC 5 under Course Policies (Part 1) to add information regarding Generative AI. (25-26 CIC 1)
   1. Faculty may define AI usage rules clearly within their course syllabi using three standardized options with templates provided: 1) AI Prohibited, 2) AI Permitted with Attribution (specify which tools and tasks), 3) AI Flexible (student choice with mandatory disclosure).
   2. Per CSUEB annual AI program review cycle, departments to report course-level policies to Academic Affairs and CAPR via Annual Program Review protocol - for curricula review and program development for CO compliance on AI metrics.
5. Student AI Literacy and Skills Development: For meeting our workforce preparation strategic goal, student preparation shall focus on developing literacy through fluency.
   1. Emphasizing workforce preparation – as part of CSUEB’s mission and CO new strategic initiative - new program and course proposals may follow Purdue University model – where AI literacy is part of GE.
   2. Emphasizing workforce preparation – as part of CSUEB’s mission and CO new strategic initiative - new program and course proposals may follow Ohio State University model – where AI fluency is integrated into general education, First Year Success Series, and field-specific coursework to ensure every Ohio State graduate in the class of 2029 will be AI fluent.
   3. To develop AI literacy thru fluency, CSUEB shall coordinate initiatives across units and supporting the development of assignments and modules that explicitly teach students how to use AI tools effectively, ethically, and critically,
6. Academic Honesty Integration:
   
   1. To assist AI literacy thru fluency development, CSUEB shall align GenAI use with updated Academic Integrity policies that clearly differentiate between acceptable assistive use and misconduct, reflecting system-wide standards (e.g., CSU or WASC guidance).
   2. To increase awareness of responsible and ethical AI use, CSUEB shall communicate academic integrity policy and updates on multiple channels.

Section III. Supporting Equity, Accessibility, and Responsible Innovation

CSUEB is committed to adopt AI tools that feel human-centered to enhance user experience and team productivity while offering trustworthy solutions with robust security guardrails.

1. Human-Centered Approach: Decisions affecting individuals must be made or reviewed by humans, and no critical academic or disciplinary decisions may be delegated solely to AI tools.
2. Equity Assurance: AI systems will be evaluated regularly to ensure they work fairly for all campus stakeholders including student groups, with mechanisms to report and rectify disparities.
3. Universal Design for Learning (UDL) Integration: Technologically innovative practices and programs must ensure that AI integration does not eliminate flexible learning options that benefit students with disabilities, English language learners, or students requiring alternative assessment formats.
4. AI as Assistive Technology: Guidelines may explicitly recognize and permit AI tools as legitimate accommodations for students with documented disabilities (e.g., for note-taking, reading assistance, or writing support), separate from general academic use policies.