OU Guidelines: Artificial Intelligence Usage

Oakland University’s AI Usage Guidelines ensure that Generative AI and LLMs are used responsibly, ethically, and transparently across academic, research, and administrative functions, with strict data protection and academic integrity requirements. AI tools are permitted to support decision-making, research, and operational efficiency but cannot make final personnel, disciplinary, or student decisions, which must remain human-driven. All AI use requires disclosure, rigorous review, and safeguard

1.0 Purpose

The following AI Usage Guidelines establish a framework for responsible, ethical, and strategic use of Generative Artificial Intelligence (GenAI) and Large Language Models (LLM) across Oakland University's academic, research, clinical, and business operations.

2.0 Guiding Principles

• Responsible Innovation: Responsibly pursuing cutting-edge AI while maintaining a critical and measured approach to its implementation.

• Ethical Deployment: Ensuring AI technologies are implemented with a fundamental respect for human values, individual rights, and societal well-being.

• Data Protection: Pursuing a comprehensive strategy to safeguard Oakland University’s sensitive data from unauthorized access, misuse, or unintended disclosure throughout the AI ecosystem.

• Academic Integrity: Preserving the fundamental values of intellectual honesty, original thought, and scholarly rigor.

• Transparency: Demanding complete and honest communication about the use, capabilities, and limitations of AI technologies across all University activities.

• Risk Mitigation: A proactive approach to identifying, assessing, and systematically reducing potential negative consequences of AI technology implementation.

• Accessibility & Equity: Ensuring AI tools are inclusive and accessible, avoiding bias and disproportionate impacts on any community or population.

• Human Oversight: All AI-assisted decision-making processes must include active human involvement and accountability.

3.0 Guidelines

3.1 Data Protection

• Data used and generated by AI and LLM systems must be handled according to the University’s Data Classification Standard and external legal and regulatory frameworks.

• Confidential \Internal data must be safeguarded from unauthorized access, upload, and misuse as described in University Policy 860. 

• Entering Confidential and Internal Data into unauthorized AI systems is a public disclosure and must follow the steps defined in Oakland University’s Cyber Incident Response Program (CIRP) per the Incident Response Process.  Note: If you intend on entering Confidential Data in an AI tool please confer with UTS first.  

• All third-party AI tools must comply with applicable data privacy laws (e.g., FERPA, HIPAA, GDPR), and procurement must verify vendor compliance.

---

3.2 Authorized Use

• GenAI and LLMs will not be solely used to make personnel, award, or disciplinary decisions.

• Decision-making may not be delegated to GenAI systems.

• The university has a rigorous review process for AI technology (Software, Applications, and Website Purchasing Checklist Request), ensuring all AI tools undergo thorough evaluation by cross-functional experts before implementation.

• All approved AI systems are subject to re-evaluation upon major updates or model changes.

---

3.3 Academic Affairs Use

Research Applications

• Please refer to AI Guidance provided by The Research Office 

Teaching, Learning, and Academic Integrity 

• Please refer to AI Guidance provided by the Center for Excellence in Teaching and Learning (CETL)

---

3.4 Research Considerations

Research Data Management & Collaborative Research

• Please refer to AI Guidance provided by The Research Office 

---

3.5 University Office Use

Finance and Administration

• AI may be used for enhancing institutional efficiency, including resource optimization, workflow improvements, predictive maintenance, and strategic planning.

• AI-generated recommendations must be reviewed and approved by appropriate human decision-makers.

Planning

• AI may support strategic planning through data modeling and predictive analytics but must not replace human judgment.

• Ongoing performance monitoring is required to detect bias, errors, or deviations from institutional values.

Human Resources

• AI may support initial candidate screening or workload analysis but may not make personnel decisions (e.g., hiring, promotion, termination).

• All HR-related uses of AI require human oversight to prevent bias and protect employee rights.

Student Affairs

• AI may enhance administrative functions such as registrar processes, financial aid communication, and service efficiency.

• AI is prohibited from making independent enrollment or disciplinary decisions, which require human review and authority.

4.0 AI Governance and Oversight

• Oakland University will establish an AI Governance Council including representatives from Academic Affairs, IT, Legal, Ethics, Research, and Student Affairs to oversee policy implementation, review incidents, and evaluate new technologies.

• These guidelines will be reviewed annually to remain current with evolving legal, technological, and educational developments.

5.0 Reporting and Compliance

• Any misuse, unintended consequences, or ethical concerns regarding AI must be reported via the AI Incident Reporting Process or through the University’s CIRP.

• Violations of these guidelines may result in disciplinary or administrative action, as outlined in applicable university policies.