AI-Generated Content Guidelines

As the availability of Large Language Models (LLMs) and generative AI tools has become widespread, North Carolina A&T State University requires the following guidelines to ensure the responsible, accessible, and ethical use of AI-generated content across all university digital properties.

These guidelines apply to text, code, images, and multimedia generated for the university website, communications, and digital platforms.

Data Security and Privacy

  • Protect Sensitive Information: Never input confidential university information, student data protected by FERPA, personally identifiable information (PII), or sensitive internal communications into public generative AI tools.
  • Approved Tools: Only utilize AI platforms that have been vetted and approved for university use when handling institutional data.

Guidelines for AI-Generated Text and Code

  • Accuracy and Quality Assurance: Review all AI-generated material for accuracy, coherence, and relevance. A human reviewer must verify that the content does not contain "hallucinations"—incorrect information presented as fact.
  • Use for Drafts and Ideation: Treat AI-generated content as a starting point for further refinement by human writers and editors. Avoid relying on AI to produce final, polished content without comprehensive review.
  • Customize and Personalize: Tailor any AI-generated content to your specific audience. Add your unique voice, expertise, and perspectives to make it engaging and relevant to the university community.
  • Accessibility and WCAG Compliance: If using AI to generate HTML structures, tables, or interactive web elements, you must strictly verify that the output meets WCAG guidelines. AI models frequently generate improperly nested headings or incorrect ARIA labels, creating significant accessibility barriers for users utilizing assistive technologies.


Guidelines for AI-Generated Images and Multimedia

  • Respect Artist Copyright: Avoid using AI-generated images that infringe upon an artist’s copyright. Understand the intellectual property implications, and ensure any AI-generated content used is properly licensed or free of restrictive copyright.
  • Accessibility and Alt Text: All AI-generated images must include descriptive Alt Text for screen readers. Visually impaired users utilizing assistive technologies must be able to understand the context of the image, and standard practice is to note if an image is artificially generated versus a real photograph of the campus.
  • Mitigate Distorted Subjects: Review AI-generated faces and bodies carefully. Avoid using distorted, unrealistic faces, or anatomically incorrect appendages that could be distracting or misleading. Consider using real photography of our campus and community for critical applications.
  • Address Garbled Text: AI-generated images often contain nonsensical text. Proofread and crop or edit any generated text to ensure it does not confuse or mislead recipients.
  • Audio and Video Transparency: Ensure that any AI-generated audio or video (including voiceovers or synthetic presenters) is clearly disclosed to the user to prevent misrepresentation.

Brand Representation and Bias Mitigation

  • Reflect Our Community: AI image generators often default to stereotypical or culturally inaccurate depictions. All generated imagery must actively align with the university's brand standards and our commitment to an accurate representation of our student body, alumni, and faculty. 

Remember that AI-generated content should always be used thoughtfully, with consideration for its impact on communication, aesthetics, and ethical standards. The A.I. Working Group Communications and Awareness team will regularly review and update these guidelines as AI technology evolves. |  In alignment with our commitment to transparency and best practices, please note that generative AI tools were utilized as a guide to assist in the research, drafting, and organization of these updated guidelines.