AI Guidelines for Faculty and Staff

AI tools are everywhere: embedded and easily accessible in web browsers, word processing programs, social media, and countless mobile applications. Students are less likely to copy and paste AI-generated content wholesale than use it in nuanced ways—brainstorm ideas, mixing AI text into their own prose, or adjusting grammar, spelling, punctuation, tone, and style (in ways marketed as “improving” their writing). Some may not even realize they are using AI, while others may deliberately disguise it using tools that “humanize” AI-generated text or insert intentional errors. Faculty may therefore miss AI use entirely, punish students disproportionately, or falsely accuse them.

Suspicion of AI use is not tantamount to academic dishonesty. “Innocent until proven guilty” is central to due process: however much one may believe “they just know” a suspect is guilty, everyone deserves fairness and reasonable doubt. Section 4.3 of the Commonwealth University Academic Integrity Policy requires investigations to follow “applicable due process requirements.” The “Supporting Documentation” section of the Student Academic Integrity Policy Violation Report Form requests relevant materials such as the assignment, exam, or syllabus. Without conclusive evidence, a charge is unlikely to be upheld. 

AI detection tools should be used with great caution. Although detection technology continues to improve, no currently available tool can reliably determine whether AI was used in the creation of a student's work. Detection scores represent probabilities rather than proof and remain susceptible to both false positives and false negatives. Faculty should not treat a detector's output as material evidence of academic misconduct or rely on it as the sole basis for initiating or resolving an academic integrity case.

If faculty choose to use AI detection tools, they should select systems that reflect current best practices and understand their documented limitations. Detection results may be considered as one piece of information within a broader review of student work, assignment context, revision history, and conversations with the student. Human judgment, due process, and the totality of the available evidence—not AI detection scores—should guide decisions regarding potential academic integrity violations.

Make your position regarding AI clear to students in your syllabus and assignments. If you allow unlimited, partial, or no use of AI, state it upfront. You may consider different AI guidelines for each assignment, depending on your learning goals. See Center for Teaching and Learning (CTL) guidelines for sample syllabus language.

Whatever other steps you take, consider fairness. For example, should students with certain disabilities or English language learners—whose writing is disproportionately flagged—be required to write by hand? Techniques like embedding invisible text or enforcing multi-stage writing can be easily circumvented. Such strategies risk approaching students from a position of distrust. Instead, faculty might:

• Have conversations with students about responsible, appropriate, ethical use of AI.
• Integrate AI into assignments, perhaps at the brainstorming or peer review stage.
• Explore AI together with students, such as showing them how to properly prompt AI tools.
• Show students both the benefits and shortcomings of AI—bias, hallucination (creating false information that may sound authentic), sycophancy (excessively praising the user), and environmental impact.
• Speak with alumni and professionals about how AI is used in their field and addresses professional standards.
• Show students how to use AI to prepare for exams (for example, by creating flashcards or sample questions).
• Take advantage of CTL, LinkedIn Learning, and other AI training opportunities.

These suggestions do not require embracing AI, but rather teaching students to use it critically and mindfully. Perhaps most importantly, consider redesigning assignments and assessments to account for changing technology. If AI output lacks qualities you value, adjust your rubric accordingly. For example, because AI is prone to hallucinating, require students to cite authoritative texts in your field. In the end, evaluate student work on its own merits using clear, consistent standards.

GenAI can be a valuable tool for both classroom instruction and workload management. When used thoughtfully, AI can assist in creating instructional materials, streamlining administrative tasks, and enhancing student engagement. Faculty might use AI to generate exam questions, study guides, and discussion prompts, with the overall goal of allowing more time for in-depth interactions with students. When followed by expert vetting, AI tools can aid in developing rubrics, drafting syllabi, brainstorming lecture outlines or innovative assignments, or adapting materials for diverse learning needs. AI should complement—not replace—faculty expertise, ensuring course materials maintain academic rigor and align with learning objectives.

Our learners are here to benefit from your expertise and deserve to know when you are using AI to design materials and activities. Just as you would expect a student to be open about their use, you should be transparent about the purposes for which you choose to use AI tools for specific tasks.

Each discipline should integrate GenAI according to its pedagogical goals and professional standards. In creative fields, AI might support idea generation or style analysis, while in STEM, it could assist with data analysis or problem-set creation. By adopting discipline-appropriate approaches, educators can help students develop the AI literacy needed for an evolving workforce.

Faculty play a key role in helping students understand the ethical dimensions of GenAI. This includes discussing how AI models are trained, examining issues of data privacy and bias, and exploring the environmental and societal impacts of automation. By integrating these topics into coursework, educators can guide students to evaluate AI outputs critically and engage with emerging technologies responsibly. When relevant, faculty should discuss the differences between various AI tools commonly used in their respective disciplines, keeping in mind that tools are changing frequently and new ones are becoming available every day.

Be mindful that some students are resistant to AI for various ethical or moral reasons. Be open to discussing these concerns with students and finding an appropriate path forward if assignments or other course activities that require AI use are problematic for students.

University staff are free to use GenAI thoughtfully to improve efficiency, communication, and service to the campus community. Given the diversity of staff roles—from administration and technology to student support and operations—AI should enhance human judgment and align with institutional values, data privacy, and equity standards. As AI technologies evolve, staff should remain informed and accountable, adopting AI where it demonstrably adds value and ensures fair, accurate outcomes.

Do not share personal, confidential, or proprietary information with AI tools. This includes student records, personnel data, unpublished research, and internal communications. Public AI systems may store or learn from input data, potentially violating University policies or privacy laws.

Example: Do not input performance evaluations or student records into AI. Use only secure, University-approved systems for sensitive material.

AI-generated content may include factual errors (“hallucinations”) or outdated information. Always verify facts, statistics, and references before using or sharing AI-generated material.

Example: If an AI provides an example or citation, double-check to make sure it is real and credible.

Respect copyright and ownership when using AI. Do not upload student work, copyrighted materials, or proprietary reports to AI tools without permission. AI may reproduce content trained from copyrighted sources; always credit original works appropriately.

Example: If AI summarizes a report or article, cite the original source in your document.

Before using an AI service, read its terms of use, privacy policy, and data-handling statements. Free tools often retain or reuse user data. Use University-approved platforms whenever possible, and avoid systems that store or share data without consent.

Be transparent with students and colleagues about when and how AI tools contribute to written or instructional materials. Include clear attribution when AI assists substantially in producing syllabi, reports, emails, or other documents.

Example: “Portions of this syllabus were drafted with assistance from OpenAI’s ChatGPT and subsequently reviewed and edited by the instructor.”

Use AI in ways consistent with the University’s values of honesty, fairness, and respect. AI should support, not replace, human expertise or critical thinking. Revise, fact-check, and ensure all final work reflects your professional judgment and authentic voice.

Example: Encourage students to engage with AI ethically: for instance, as a brainstorming aid, not as a shortcut to complete assignments

Effective AI use depends on clear, specific instructions. Provide context, tone, and purpose in your prompts to guide high-quality output. Consider using a framework (e.g., TRACI: Task, Role, Audience, Create, and Intent) to achieve more effective results.

Example: “Draft a concise, professional memo to staff announcing a campus event, with a friendly and inclusive tone.”

AI can accelerate writing tasks but does not replace human review. Faculty and staff are responsible for editing, approving, and ensuring accessibility in all materials. AI-assisted content should meet ADA and WCAG standards for readability and inclusivity.

Example: Review AI-generated materials to confirm they are readable by screen readers and use plain, inclusive language.

An AI-generated image is visual content where the primary composition and appearance or key visual elements are determined by AI software rather than human manual effort. This content is typically created by an AI program based on a user's text prompt, reference images, or data input, making it distinct from traditional photography or drawing. It may also include images that incorporate AI-generated components alongside traditional photography or illustration.

• When it’s an abstract content
  • Example: Illustrating abstract research concepts (e.g., quantum physics).
  • Why It’s Okay: AI images can visually represent ideas that don't exist in the real world.
     
• When the disclosure is clear
  • Example: Including a caption to declare: Image generated with AI.
  • Why It’s Okay: Transparency can help to maintain credibility.
     
• When having a real image isn’t possible or is impractical
  • Example: Visualizing ancient history or possible future scenarios.
  • Why It’s Okay: AI offers creative flexibility without misrepresenting reality.
     
• When artistic projects use AI as a predetermined or prequalified focus
  • Example: Digital arts programs showcasing AI-generated art created by students or faculty.
  • Why It’s Okay: Highlights creativity and innovation in an appropriate academic context.
     
• When used as a supplement, not a substitute, for a primary focus, like a hero graphic
  • Example: Background visuals behind real student or faculty testimonials.
  • Why It’s Okay: It can enhance storytelling without replacing real people or events.

• When the image contains or recreates an official university logo or mark
  • Example: An AI-generated version of a campus spirit mark (e.g., Bloomsburg’s “B dog U” logo) on a submitted T-shirt design or webpage submission.
  • Why It’s Not Okay: AI tools can distort, alter, or inaccurately reproduce official trademarks, creating compliance and branding issues. These assets are official symbols of the institution and must remain consistent, accurate, and protected. University logos and marks must always come directly from approved brand assets. AI-generated or modified versions should not be used unless explicit authorization has been granted by Strategic Communications or the president’s office.
     
• When depicting real people, events, or campus life
  • Example: Using AI-generated students to represent your students, faculty, or staff within our website directory, on a general webpage, or admissions materials.
  • Why It’s Not Okay: This can mislead users, misrepresent individuals, and undermine institutional trust. Directory images must be authentic photographs of the actual person, not AI-generated substitutes or approximations.
     
• When disclosure is unclear
  • Example: Publishing AI images without noting they are not real photographs.
  • Why It’s Not Okay: Lack of transparency can be perceived as manipulation or misinformation.
     
• When sensitive or scientific contexts are necessary
  • Example: Medical or scientific research illustrations without proper disclaimers.
  • Why It’s Not Okay: Could lead to misinterpretation or the spread of inaccurate information.
     
• When it replaces authentic representation
  • Example: Using AI art instead of featuring work by actual students or faculty.
  • Why It’s Not Okay: It may undercut actual contributions and authenticity.
     
• When depicting diversity or identity
  • Example: AI-generated images of diverse student groups for DEI pages.
  • Why It’s Not Okay: Misrepresents actual community demographics and can be exploitative.

• Always disclose that the image is AI-generated.
• Use alt-text and image descriptions that clarify the image is not real.
• Prioritize authenticity: Use real photos of students, faculty, staff, and facilities whenever possible.
• Follow applicable state and federal laws along with institutional policies. 
• Respect copyright and ethical boundaries when using AI tools trained on artist or photographer work.

• When demonstrating technology or instructional simulations 
• When exploring AI media for artistic or research purposes 
• When presenting clearly labeled experimental or illustrative content 
• When conducting brainstorming sessions 
• When prototyping and storyboarding 
• When drafting script (with full disclosure) 

• When recreating or simulating real people without disclosure or consent
  • Example: Generating a video of a university leader speaking words they did not say. 
  • Why It’s Not Okay: This can mislead audiences, damage reputations, and undermine institutional trust.
     
• When creating synthetic video of students, faculty, or campus life meant to appear authentic 
  • Example: AI-generated “student testimonial” videos.
  • Why It’s Not Okay: Misrepresents the real university experience.
     
• When using AI to manipulate existing video of university members in misleading ways
  • Example: Altering a recorded speech using AI voice or face replacement.
  • Why It’s Not Okay: Raises ethical, legal, and reputational risks.

• AI tools may create copyright and/or ownership issues:
  • If the video is partially or fully created from an existing work that is copyrighted
  • If it’s unclear about the ownership rights of the video

The user must be aware of the AI tools terms and conditions and licensing guidelines.

AI-generated or modified video and audio should be clearly disclosed and reviewed by appropriate communications or leadership offices prior to publication.

Vague policies on AI use can cause confusion and inequity. Instructors should explicitly state when, how, and to what extent AI may be used—for example, whether it may assist with brainstorming, grammar checking, or citation formatting. Transparent guidelines promote ethical learning habits and help students develop critical digital literacy.

AI can support administrative efficiency, but it must not replace faculty judgment in evaluating student work. Automated grading of complex or creative assignments risks misinterpreting nuance, reinforcing bias, and eroding trust. AI tools may assist with objective assessments or pattern detection, but faculty remain responsible for final evaluations and individualized feedback.

Student work is protected intellectual property and should not be uploaded to AI platforms without express permission. Likewise, confidential research data or personal information must never be shared with unvetted systems. Only public or explicitly authorized materials should be processed by AI tools.

Faculty and staff should disclose when AI meaningfully contributes to course design, grading, or feedback. Assignment guidelines should require students to document their own AI use through citations or reflective notes. Both educators and students must remain aware of potential biases in AI systems and discuss how these biases may shape outcomes.

The Center for Teaching and Learning (CTL) supports faculty and staff in exploring, implementing, and reflecting on responsible uses of GenAI in teaching, research, and professional practice. CTL provides training and resources to help the university community stay informed, confident, and creative in this evolving landscape. These activities are not an endorsement to use any particular AI tool, but are offered to help educate faculty and generate discussion about appropriate uses of AI in our varied fields.

The CTL Brightspace serves as the central repository for AI-related materials, including curated articles, videos, and on-demand courses from LinkedIn Learning and the ACUE Learning Lab. Faculty-led workshops are recorded and made available for flexible, just-in-time learning.

To foster continuous growth and community, CTL hosts regular conversations on emerging AI topics, classroom applications, and ethical considerations. These sessions invite faculty to share strategies, discuss challenges, and showcase innovative practices across Commonwealth University. By cultivating this culture of collaboration, CTL ensures that faculty and staff remain supported and empowered as they adapt to technologies shaping the future of higher education.