Generative AI models are tools that use large language models to generate text responses to user prompts. They identify patterns in language based on training data and produce responses that sound conversational, but they do not understand content or verify facts the way a human does. Many generative AI systems have been working on improvements in the areas of image generation and coding, among others. All assistance with chat, text, and coding can be put into the chat box of the three AI tools available through the university.

Yes. Three AI tools, Copilot, ChatGPT, and Gemini, are available for faculty use with an approved license. Faculty should use these tools in accordance with the University of Houston AI Guidelines responsibly and safely. See https://uhsystem.edu/offices/information-security/resources/artificial-intelligence/index.php

• Drafting and refining course materials and plans
• Generating ideas for assignments, assessments, or discussions
• Reviewing and improving the clarity of written content
• Supporting coding, data exploration, or technical problem solving

Note: All AI-generated outputs should be reviewed and edited by faculty before use.

Faculty should not:

• Upload confidential, sensitive, or student-identifiable information. See: Data Classification and Protection | University of Houston System
• Rely on AI-generated content without review or verification
• Use AI tools to replace professional judgment for grading, student learning, and evaluation

All generative AI tools may produce inaccurate, incomplete, or biased responses, often referred to as hallucinations. Faculty should treat AI output as draft/exploratory content and independently verify accuracy before use.

AI licenses obtained through the University of Houston using your “username@cougarnet.uh.edu” email address for access are secure, keeping individual chats and answers within the University systems ecosystem. However, faculty should avoid the use of confidential and/or sensitive information to maintain confidentiality and data privacy. See: Data Classification and Protection | University of Houston System

AI may assist in drafting rubrics, reviewing grammar and style, or generating feedback templates, but it should not be used as the method for grading or evaluating student submissions. Final grading decisions must be made by the instructor.

Student use of generative AI is a course-level decision. Faculty are encouraged to clearly communicate expectations in their syllabi regarding whether, when, and how AI tools may be used, if at all, to comply with university academic integrity policies. For more on potential concerns with AI use in writing-heavy courses, see: Cornell Research, CHI Conference

Faculty should:

• Clearly define acceptable and unacceptable AI use in course syllabi and in class with students. See sample AI Syllabus Statements here: Sample AI Syllabus Statements: 2024FE FDIS1000 203235F - Instruction@UH Tutorials
• Design assignments that emphasize critical thinking and processes that don’t use AI
• Focus on transparency and responsible use rather than detection alone