Responsible use of Artificial Intelligence (AI) means applying AI tools in a way that is ethical, transparent, secure, and aligned with the University of Guam’s values and mission.

At UOG, this includes:

1. Ensuring human oversight and accountability for all AI-assisted work.
2. Protecting confidential, sensitive, and personal data.
3. Maintaining accuracy, integrity, and the institutional voice.
4. Using AI to enhance efficiency and creativity without replacing human judgment.
5. Being transparent when AI plays a significant role in creating institutional
6. content or informing administrative
7. Continuously evaluating and adapting AI use to meet evolving best practices, regulations, and campus needs.
8. Critically reviewing AI-generated outputs for bias, inaccuracies, or incomplete context, particularly in institutional decision-support contexts.

To provide practical, UOG-wide principles and standards based on emerging best practices for the responsible use of AI tools in administrative and operational functions.

This guidance applies to all UOG employees when performing administrative, service, or operational functions on behalf of the University.

Administrative and operational functions include, but are not limited to:

• Institutional communications
• Reporting and documentation
• Program administration
• Committee or governance work
• Operational planning and management

This guidance does not regulate classroom instruction, scholarly research, peer-reviewed publications, creative works, or other activities protected under academic freedom.

Nothing in this document supersedes collective bargaining agreements, shared governance responsibilities, or established academic policies.

Use of AI tools is not mandated. These guidelines provide standards for responsible use when such tools are utilized.

AI is a tool, not a decision-maker. It may support drafting, analysis, or administrative efficiency, but it shall not replace human judgment, professional expertise, or UOG’s institutional voice.

AI tools may not be used to automate final institutional decisions without appropriate human review and policy authorization.

ALL AI-assisted content intended for public, external, or institutional release must be reviewed and approved by the appropriate administrator, director, supervisor, or designee.

AI refers to systems and tools that perform tasks typically requiring human intelligence — including, but not limited to language generation, image creation, pattern recognition, summarization, and decision support.

Faculty, staff, and administrators use many different AI tools at UOG including: ChatGPT (OpenAI), Claude, Perplexity, Grok, Microsoft Copilot, Google Gemini, Zendesk Chat, Canva, Grammarly, DALL·E, Nano Banana,  Adobe Firefly, Midjourney, and others.

The inclusion of specific tools in this document does not constitute institutional endorsement or procurement approval.

AI tools can enhance productivity, clarity, and efficiency — but should be used responsibly, ethically, and with human oversight.

1. Human accountability is required. All AI-assisted content must be reviewed by a human before publication, dissemination, or use in institutional decision-making.
2. Data privacy must be safeguarded. Protected data (student or employee information, PII, FERPA-protected records, HIPAA-protected health data, or sensitive financial data) must not be entered into public AI tools.
3. Enterprise vs. Public Tools Distinction. Employees should distinguish between enterprise-approved AI systems operating within secure institutional environments and public, consumer-facing AI platforms. Public AI tools should be treated as third-party services unless covered by an approved data protection agreement.
4. Quality, authenticity, and brand integrity must be upheld. AI should enhance, not replace, human judgment or UOG’s institutional standards.
5. Bias awareness and fairness. AI outputs may reflect embedded biases or incomplete training data. Users are responsible for critically reviewing AI-generated content to mitigate bias and ensure fairness.
6. Ethical alignment. AI use should align with UOG’s mission, values, and public trust.

1. Summarizing reports, meeting notes, and documents.
2. Drafting content for websites, FAQs, or internal
3. Drafting non-technical portions of grant proposals or reports (with appropriate oversight).
4. Assisting with drafting internal and external communications.
5. Assisting with qualitative analysis (e.g., summarizing survey comments).
6. Generating content variants for presentations or multi-channel communication.

1. Inputting protected or sensitive data into public AI tools.
2. Using AI as the sole or determinative basis for hiring, tenure, promotion, evaluation, admissions, or disciplinary decisions. All such decisions must be made by qualified human reviewers in accordance with applicable policies and collective bargaining agreements.
3. Generating legal or contractual documents without review by UOG’s General Counsel.
4. Publishing AI-generated institutional content externally without human editing and approval.
5. Fabricating research findings, citations, or references.
6. Using AI to impersonate individuals or misrepresent institutional positions.

AI-generated visuals may be used for illustrative, conceptual, or draft purposes. However:

• AI visuals should not impersonate real individuals without consent.
• AI should not be used to fabricate documentation of real UOG events or achievements.
• AI may not be used to recreate or alter official UOG seals, logos, and protected brandmarks and wordmarks.
• When AI-generated visuals could reasonably be mistaken for authentic photography or documentation, disclosure (e.g., “AI-generated image”) is encouraged.

Human review and supervisory approval are required before publishing AI-generated visuals in official materials for public dissemination.

Custom AI assistants and AI agents may be used for administrative purposes provided that:

• No sensitive data is used without proper authorization.
• Outputs are reviewed by a human before institutional use.
• They are not used to bypass approval processes.
• Large-scale or sustained AI integrations are coordinated with the Office of Information Technology (OIT) to assess security, privacy, and contractual considerations.

While routine disclosure of AI assistance is not required, employees must not misrepresent or deny the use of AI tools if directly asked.

Disclosure is appropriate when AI tools have materially influenced the substance, tone, or conclusions of a product. For the purposes of this guidance, “material influence” refers to instances where AI-generated content meaningfully shapes the message beyond routine drafting support (e.g., structuring arguments, framing positions, or generating substantive portions of the final output).

Routine use of AI for drafting, editing, formatting, or administrative correspondence does not require disclosure. When uncertainty exists, employees are encouraged to consult supervisors or err on the side of transparency, where doing so supports institutional credibility.

Routine internal drafts or brainstorming may involve AI without formal disclosure. However, any content intended for official or public release must undergo human review and approval.

AI systems may support analysis but may not independently determine institutional outcomes.

All non-public institutional data should be treated as sensitive.

If uncertain whether information may be entered into an AI system, employees should consult OIT or their supervisor.

Units are encouraged to use AI tools thoughtfully and proportionately, recognizing both their benefits and broader resource implications.

The University is currently developing basic AI training for all employees and students and encourages employees to build AI literacy consistent with institutional values and professional standards.

Administrative units may designate a point of contact to coordinate AI training and safe-use practices as appropriate.

Faculty and researchers are encouraged to explore AI responsibly within their disciplines under academic integrity standards and governance structures.

AI technologies continue to evolve. UOG may explore future applications such as AI-assisted advising, operational analytics, or workflow automation, subject to appropriate governance and human oversight.

AI enhances institutional capacity — but human judgment, accountability, and integrity remain central to UOG’s mission.