
Generative AI
Overview
These guidelines are for marketing and communications and not intended to apply to other areas of the university.
Generative artificial intelligence (AI) services, which generate text, images or other media in response to human-generated prompts, are rapidly growing and evolving. Many generative AI applications carry significant potential to support human innovation across a wide range of marketing communications disciplines. These include AI tools used to generate content, such as images, text, music/audio, software development, video, web and app development, and design/illustration.
University Communications encourages exploration of generative AI services with the following guiding principles in mind:
AI Should Augment, Not Replace, Humans
Ensure humans review AI results for accuracy of information and potential bias.
Be transparent. Use Citations and Check Them for Accuracy
The use of material taken from any source—whether directly quoted, paraphrased or otherwise adapted—should be attributed to that source. This includes AI.
Any and all AI results should use traditional research methods to fact check, and you must accurately confirm and document correct sources and accurately cite all AI-generated text, images or videos.
Be inclusive. Monitor for and Reduce Bias
AI "learns" using data-based assumptions from a wide data set. AI will respond in a similar fashion to what was in the prompt. Including bias in your prompts will generate biased content. All materials generated using AI tools must be carefully reviewed, approved, edited and overseen by a human author.
Be safe. Protect your data
Any prompt or materials provided to AI platforms may be stored and used for subsequent responses to other users. Refrain from using any personally identifiable or confidential information when working with any AI platform.
Sharing or providing any trademarked university logos, images, videos or other proprietary files, code, text or institution data is strictly prohibited.
Be respectful. Respect Individual Privacy
Using personally identifiable information (PII) in an AI prompt puts private information at risk of theft, and the person at an increased risk of identity theft and having their PII released publicly. Never provide personal information or employee information (benefits, salary, performance plans, etc.) to any AI platform.
Appropriate Marketing and Communications Outcomes and Uses
The use of generative AI tools for internal and external communications and marketing is subject to specific guidelines that align with our brand values, ethos, ethical standards and legal obligations. These tools may be used only to assist with creative work; they may not be used on their own or to replace the role or importance of human creators. As with all work (with or without the use of AI), every effort must be made to ensure accuracy, eliminate bias, and uphold privacy and confidentiality.
The following list of approved and prohibited tools is not exhaustive. If you have questions about the approved use of a specific application, please reach out to ur@uccs.edu.
Example: OpenAI's ChatGPT, Microsoft Copilot, Grammarly, Otter.ai (transcription)
Acceptable Use:
- Brainstorm, research a topic for a story, other text-based content
- Draft an outline for a story, other text-based content
- Summarize, edit text/interviews or stories
- Check and correct grammar
- Suggest edits to improve writing quality (rephrasing, thesaurus)
CAUTION regarding data security/privacy: All text entered into large language models like ChatGPT becomes part of the model. Do not place confidential or highly confidential information in the application.
Unacceptable Use:
- AI-generated content that:
- is presented as original work
- is unchecked for accuracy, bias and/or confidentiality
- violates brand, ethical and/or legal standards
- is not identified as AI-generated or assisted content
Standard: Enhance creativity and productivity but maintain originality. Tools should assist in productivity and skill development, not replace effort.
Example: Adobe Creative Suite for editing, Adobe Firefly for image creation for creative ideation only.
Acceptable Use: Editing of original images (e.g., color, composition, exposure, etc.) and the removal—not addition—of objects in an image for which we own the IP. Creation of images for ideation, enhancing the creative process that will not be used in any marcom content (internally or externally).
Unacceptable Use: At this time, the use of AI-generated images, music, audio and video—even if used to add objects to a university-owned asset—and including use for social media, is prohibited in any university communications. AI for audio and visual content can be used only to help brainstorm ideas.
Standard: Foster creativity within the bounds of intellectual property laws.
Social Media & Website/Digital Content Creation & Management
Examples: OpenAI's ChatGPT, Sprout Social AI Assist, Google Ads AI, SEMrush
Acceptable Use:
CAUTION regarding data security/privacy: All text entered into large language models like ChatGPT becomes part of the model. Do not place confidential or highly confidential information in the application.
Unacceptable Use:
Standard: Enhance creativity and productivity but maintain originality. Tools should assist in productivity and skill development, not replace effort.