Staff - Ethical AI Usage - NMU AI Literacy Initiative

• Step 1: Familiarize yourself with relevant data protection regulations like FERPA and GDPR.
• Step 2: Ensure that AI tools used at the university are secure and ethically aligned with university values.
• Step 3: Participate in AI ethics training sessions to stay informed about best practices for ethical AI use.
• Step 4: Implement regular audits and security checks to ensure AI systems remain secure and free from bias.

As universities increasingly adopt AI tools and technologies, it is crucial to address ethical concerns and ensure data security. Staff play an essential role in maintaining ethical standards in the use of AI and protecting sensitive information from misuse. This page provides guidance on how university staff can responsibly use AI tools, maintain privacy and data security, and follow best practices to protect institutional and personal data.

1. Understanding Ethical AI Use

Ethical considerations in AI use are fundamental to ensuring that AI tools benefit everyone equitably, without causing harm or bias. Staff should be aware of potential ethical issues and work to ensure that AI use aligns with university policies:

• Transparency and Accountability:
  • Staff must be transparent about how AI is being used in university processes and ensure that students, faculty, and other stakeholders are informed of its presence. This includes providing clear guidelines for the ethical use of AI in administrative tasks, research, and teaching.
  • AI Accountability: Staff must also hold AI tools and their use accountable. This includes reviewing AI-generated decisions to ensure fairness, identifying and addressing potential biases, and ensuring that the AI is used to augment human decision-making rather than replace it.
• Bias and Fairness:
  • AI algorithms can unintentionally perpetuate biases based on data they are trained on. University staff should ensure that AI tools are regularly tested for fairness and adjusted to eliminate biases in admissions, hiring, grading, or any other use.
  • Diversity and Inclusion: AI tools should be designed and used in ways that promote diversity and inclusion, considering factors such as gender, race, socioeconomic status, and disability to prevent AI from reinforcing existing inequalities.

2. Protecting Personal and Institutional Data

Data security is one of the most critical aspects of ethical AI use. Ensuring that sensitive personal and institutional data is kept secure is a key responsibility of university staff:

• Data Privacy and Compliance:
  • University staff must be familiar with data privacy laws such as FERPA (Family Educational Rights and Privacy Act) and GDPR (General Data Protection Regulation), which govern the collection, use, and sharing of personal data. Ensuring compliance with these regulations is essential for maintaining trust and safeguarding students' and staff members' privacy.
  • Anonymization and Encryption:
    • Personal and sensitive data should be anonymized whenever possible, ensuring that AI tools cannot trace data back to individuals. Encryption techniques should also be used to secure data during transmission and storage.
• Data Access Control:
  • University staff should implement access controls to ensure that only authorized personnel can access sensitive data, both for AI training and everyday operations. Strong authentication methods, such as two-factor authentication (2FA), should be adopted to protect sensitive systems.
  • Staff should regularly audit data access logs to ensure compliance with access controls and identify any unauthorized data access attempts.

3. AI Training Data Integrity

The integrity of training data directly impacts the quality and fairness of AI outputs. Ensuring that AI systems use reliable and ethical data is a key responsibility for staff:

• Data Sourcing and Integrity:
  • Staff should ensure that training data used in AI systems is obtained from reputable sources and is free from biases. This includes avoiding data that may reinforce stereotypes or provide unfair advantages to certain groups.
  • Data Quality:
    • Ensure that the data used to train AI tools is of high quality—accurate, up-to-date, and relevant to the intended use. Low-quality data may lead to poor decision-making, misclassifications, or other issues that could harm students, faculty, or the institution.

4. Secure Deployment and Use of AI Tools

Implementing AI tools securely is essential for protecting data and maintaining the trust of university stakeholders:

• Secure AI Platforms:
  • Staff should work with trusted vendors that prioritize security when providing AI solutions. Before deploying any AI tool, staff should evaluate its security features, such as encryption standards and vulnerability testing.
  • University IT departments should be involved in the selection and review of AI tools to ensure that they meet institutional security standards and are compatible with the university’s existing systems.
• AI Usage Monitoring:
  • Regularly monitor AI tools and their outputs to detect any anomalies or signs of security breaches. Staff should also track the performance of AI systems to ensure they are functioning as intended and that no unintended consequences arise from their use.

5. Mitigating Risks and Managing AI-Related Incidents

Even with the best precautions, risks related to AI use can arise. University staff should be prepared to handle these risks and manage any incidents effectively:

• Incident Response Protocols:
  • In the event of a data breach or other security incident involving AI tools, staff must follow institutional incident response protocols to mitigate damage. This may include notifying affected individuals, investigating the cause of the incident, and reporting to relevant authorities as required by law.
  • AI Ethics Committees:
    • Universities should establish ethics committees or working groups to oversee AI usage, assess ethical implications, and address any issues that arise from the deployment of AI tools on campus.
• Bias Audits:
  • Staff should periodically audit AI systems for biases, ensuring that the tools continue to operate fairly as they evolve. AI audits should be conducted regularly, and the results should be shared with relevant stakeholders to ensure transparency.

6. Training and Educating Staff on Ethical AI Use

Continuous education and training are key to fostering ethical AI practices across the university:

• AI Ethics Training:
  • Universities should provide AI ethics training for staff, covering topics such as fairness, transparency, data privacy, and security. This ensures that staff are aware of their responsibilities and understand how to use AI tools in a way that aligns with institutional values.
  • Cross-Department Collaboration:
    • Encouraging collaboration between departments such as IT, research, and administration can help ensure that AI tools are used responsibly across all areas of the university.
• Resources for Ongoing Learning:
  • AI Now Institute: An interdisciplinary research center focused on the social implications of artificial intelligence.
  • Ethics Guidelines for Trustworthy AI: European Commission guidelines on ensuring AI is ethical and trustworthy.
  • OpenAI Ethics: Provides insight into ethical AI development practices from OpenAI, one of the leading AI research organizations.