UNLV administrators understand the value of AI tools in optimizing workflows. To preserve the integrity of UNLV communications, these guidelines provide a framework for their responsible use. As AI evolves, the guidelines will be reviewed regularly and we recommend communicators reference this document often. Always employ human judgement when using AI outputs and be aware of misuse or misapplication.

These guidelines propose the following measures:

Disclosure of AI Usage

All AI-generated content posted on UNLV’s social media accounts must follow each platform's disclosure policies.

AI usage on UNLV’s owned platforms (websites, News Center) should follow the written, image, and audio guidelines listed in this document. Clearly acknowledge to editors if AI is used to generate communications content. If you are concerned, discuss with an editor.

Privacy Guidelines

When using AI tools like ChatGPT, sensitive and personally identifiable information must not be included in the prompts. Use only the minimum necessary amount of data in your inputs. Ensure that information complies with HIPAA and FERPA, as well as UNLV’s cybersecurity and data protection policies.

This includes not using NSHE ID numbers, employee ID numbers, intellectual property, and embargoed information (such as research studies and draft reports) in your prompts.

Follow the best practices outlined by IT regarding permissions, AI monitoring, and data security compliance.

Generative AI Writing Guidelines

Generative AI cannot replace human writing but it can be used as an editorial tool in the content production process. It should not be used to create “original” writing for articles, web content, or social media posts.

Communicators must always manage and be held accountable for their content. All AI output used in communications content must be reviewed and fact checked by the writer and by editors.

Appropriate uses of AI tools include:

  • Creating content outlines based on information vetted by the writer
  • Drafting content summaries of the completed article for consideration by newsletter editors and social media managers
  • Checking for grammar, spelling, and formatting to AP Style
  • Optimizing content by identifying strategic SEO keywords to include in titles, headings, and body text
  • Drafting transcripts of recorded interviews
  • Improving clarity and organization

Inappropriate uses of AI tools include:

  • Creation of significant portions of text
  • As the source for factual information
  • Fabrication of quotes

Image and Audio Guidelines

Many image and video softwares now employ AI models to assist with editing tasks, which is generally an acceptable application of an AI tool. These editing AI tools should be used carefully to maintain authenticity and integrity of our content.

AI should not be used to fully generate still or video images, audio or voiceovers; to substantively alter the content of the image or voice; or to mislead the viewer about the authenticity of the visual.

For use with communications content, AI can be used to:

  • Remove backgrounds in headshots or apply creative elements that don’t alter meaning (with adequate disclosure in the metadata)
  • Assist with tasks such as color grading or noise reduction
  • Make collage images from existing photography
  • To create placeholder visuals or audio to assist in production
  • To make make touch ups of temporary blemishes, as can be done with existing digital photography tools

Limited exception include:

  • To create a conceptual image or sound to go complement or demonstrate stories about AI-related teaching or research

Avoid using AI-generated stock assets (images, music) in your communication products.

Many AI models are trained on their prompts. Do not include UNLV brand assets in your prompts to prevent them from being added to AI training models. Content created with AI is not protected under copyright law. Do not use AI to generate UNLV brand assets.

Other Considerations

If you didn’t bother writing it, why should your audience bother reading it?

We recommend AI usage as a process, not a product. Always take care with your communications. If AI-generated content could be seen as misleading, it’s best not to publish or post it. Avoid diluting the UNLV brand. Remember your role as a communicator to maintain your responsibility of protecting and promoting UNLV.

AI Trends on Social Media

The conversation around AI-generated content to take part in a social media trend is nuanced. Participating in these trends can lead to criticism about not investing in local or student artists as well as the ongoing debate about the cost of using AI-generated content.

When AI trends circulate, we recommend finding a way to create original content based off of the trend rather than using AI or not participating in it at all. (For example, a recent AI content trend was to generate an action figure representing your university. In this case, we'd recommend working with a graphic designer to create the art rather than an AI engine.)

Related Information

AI Checkers

 

Contacts