Ethical Use of AI

AI in teaching and instructional contexts can enhance educational delivery by automating administrative tasks, improving learning methods and experiences, and supporting instructional strategies. It can aid in curriculum development, analyzing educational trends and student performance data, and helping instructors refine their teaching methods and materials. Overall, AI often boosts efficiency, enriches the learning experience, and enables more effective, data-informed instruction. To ensure ethical AI use in education, UNT Health faculty should focus on transparency by clearly explaining AI tools’ roles and practices to students while ensuring program quality is maintained. Bias and fairness must be continuously monitored to prevent perpetuating inequalities, and informed consent should be obtained for data collection. Accountability involves maintaining human oversight and addressing concerns, while regular reviews and adherence to ethical standards will ensure AI tools are effective and equitable. Accessibility for all students, including those with disabilities, must also be ensured. As a university, UNT Health and its faculty are responsible for the quality, integrity and review of its programs.

AI has the opportunity to transform workflows at UNT Health. Recognizing the power that AI tools bring, this guidance aims to protect the integrity of UNT Health’s programs, services, and reputation. Utilizing AI outside these guidelines may expose you and the institution to additional liability. Examples of potential liability related to AI include:

• Reputational Damage: An individual’s improper use of AI can harm the reputations of institutions (UNT Health) as well as the individual. This means that not only could UNT Health’s credibility be damaged, but individual instructors’ careers could be damaged as well.
• Intellectual Property Infringement: Using AI-generated content that lacks proper sourcing and citation may lead to violations of intellectual property rights and expose private data, including students’ information, to unauthorized sources.

Instructors have the ability to impact the reputation of the institution and thus have responsibility to uphold the values and academic standards of UNT Health. This guidance exists to help mitigate risk and promote a culture of integrity related to the ethical use of artificial intelligence.

Accordion Body Content

1. Consider your role. Guidance exists to support different roles at UNT Health (e.g. for students, instructors, staff, researchers). When operating in roles beyond an instructor (e.g. as a researcher), other guidance may be appropriate and helpful to consult.

In Student Coursework

2. Determining and communicating allowed tools and resources: Instructors are responsible for determining which AI tools and resources that students are permitted to use for course assignments and assessments. It is crucial for instructors to clearly communicate these guidelines to students to ensure transparency and consistency.

To maintain clarity and fairness:

• Authority: Instructors have the authority to specify the tools and resources that can be used for assignment or assessment.
• Communication: These guidelines should be explicitly outlined in the course syllabus or included in the assignment instructions.
• Consistency: All students must be informed of the allowed tools and resources at the time the assignment or assessment is issued to prevent confusion and maintain academic integrity.
• Integrity: At UNT Health, unauthorized use of AI by students to complete assignments is considered cheating based on the Student Code of Conduct and Civility (UNT Health Policy 7.105). Note that this specifically means that their use must be unauthorized, which requires communication and clear distinction from faculty as to which uses are and are not

3. Using generative AI tools in assignments: Instructors should clearly state their intentions regarding the use of generative AI tools by students. To ensure that students learn effectively and adhere to their expectations. It is essential for students to understand the boundaries set by the instructor for AI tool usage.

Recommendations:

• Disclosure: Instructors should explicitly outline the appropriateness or inappropriateness of generative AI tool use in their course syllabus.
• Assignment Instructions: Clearly indicate any restrictions or allowances for AI tools in individual assignment instructions.
• Consistency: Ensure that students are aware of these guidelines and expectations from the outset to support their learning process and maintain academic integrity.

 

4. Managing AI tools in coursework: It is crucial for instructors to understand several forms of which students might use AI tools in relation to the coursework assigned. To integrate AI into the learning process or acknowledge its potential use by students:

• Encourage AI Use: If you intend for students to use AI tools for an assignment, assessment, or learning activity, evaluate how AI use aligns with the designed learning outcomes. Ensure that the use of AI supports students in achieving these outcomes rather than hindering their understanding.
• Restricting AI Use: If you prefer students not to use AI tools, consider how the task is assessed. Design assignments in a way that minimizes the effectiveness of AI tools and promotes the intended learning objectives. This may involve creating tasks that require critical thinking, personal reflection, or unique problem-solving skills that AI tools cannot easily replicate. Additionally, live in-person assessments, and assessments that use established tools for creating a proctored environment are valuable in restricting AI use.

5. Using AI in developing educational tools. AI tools offer valuable support in enhancing educational content by assisting with the ideation and creation of lesson plans, questions, scenarios, images, and other course components. Utilizing AI can save time and reduce the creative burden associated with developing these elements. However, it is important to maintain accountability and ensure the quality and accuracy of the course content.

Recommendations:

• Responsibility for Content: Regardless of the AI tools used, you are responsible for the quality, accuracy, and appropriateness of the course content. Ensure that all materials meet the educational standards and learning objectives of your course.
• Disclosure of AI Use: It is recommended to disclose the use of AI tools in the development of educational materials to maintain transparency. This can be achieved by:
  • Informing students: Instructors should mention in the syllabus or course documentation that AI tools have been used to assist in creating course components. This helps students understand the resources and technologies involved in their learning.
  • Providing context: Explain how AI tools have contributed to the content, if relevant, provide the students insight into the development process and its impact on their learning experience.

6. Using AI in assessing student work. When assessing student work, adherence to FERPA regulations is essential to protect student privacy. Uploading a student’s entire assignment to third-party websites, including AI tools or AI detectors, may violate their rights to privacy. Only third-party entities with which the university has specifically entered into a contract, and which have agreed to maintain the privacy of our student data are permitted for this function.

Recommendations:

• FERPA Compliance: Ensure that any use of AI tools in assessing student work complies with FERPA guidelines, which govern the handling and confidentiality of student records. Avoid uploading entire assignments to third-party platforms that could compromise student privacy.
• Disclosure of AI Use: If AI tools are used in the assessment process, it is best practice to disclose this to students. This includes:
  • Informing Students: Clearly communicate in the syllabus or assessment instructions if AI tools will be used to evaluate their work. This transparency helps students understand how their submissions will be assessed and the role of AI in the evaluation process.
  • Providing Details: Outline how AI tools will be used, what data will be analyzed, and the measures in place to protect student privacy.
  • If an instructor does not plan to disclose the use of AI in assessing student work, they should refrain from using AI tools for this purpose.
• Accreditation: It is a condition of university accreditation that UNT Health provides qualified experts to teach and assess our students. Therefore, the use of AI tools alone to evaluate student work is not permissible. In each instance, if AI is used to assist in an instructor’s or TA’s grading or assessment workflow, the AI generated feedback must be reviewed by a qualified expert for accuracy and appropriateness.

AI in teaching and instructional contexts enhances educational delivery by automating administrative tasks, personalizing learning experiences, improving learning methods, and supporting instructional strategies. AI can assist faculty in improving the quality of their courses through the analysis of student achievement and progression. Overall, AI may boost efficiency, enrich the learning experience, and enable more effective, data-informed instruction.

Additionally, AI can support students by facilitating brainstorming, developing initial concepts for projects, and aiding in the drafting of academic materials. To ensure the ethical use of AI in education, UNT Health faculty should prioritize transparency by clearly explaining the roles and practices of AI tools to students. Continuous monitoring for bias and fairness is essential to prevent perpetuating inequalities, and informed consent must be obtained for data collection. Accountability requires maintaining human oversight and addressing any concerns, while regular reviews and adherence to ethical standards will help ensure that AI tools are both effective and equitable. Finally, it is crucial to ensure accessibility for all students, including those with disabilities.

AI has the opportunity to transform workflows at UNT Health and beyond. In recognition of the power that AI tools have, this guidance aims to protect students and UNT Health from AI misuse and the potential consequences. Examples of potential liability related to AI include:

• Reputational Damage: An individual’s improper use of AI can harm the reputations of institutions (UNT Health) as well as the individual. This means that not only could UNT Health’s credibility be damaged, but individual instructors’ careers could be damaged as well.
• Intellectual Property Infringement – Using content provided by AI that should be properly sourced and cited AND exposing private data, including the data of students to unauthorized sources.

This guidance exists to help mitigate risk and promote a culture of integrity related to the ethical use of artificial intelligence.

1. Consider your role. Guidance exists to support different roles at UNT Health (students, instructors, staff, researchers). When operating in other roles (e.g. a TA, a researcher), other guidance may be appropriate and helpful to consult.

 

2. You are responsible for your own work. Regardless of which resources a student chooses to use, students are responsible for the quality and correctness of the coursework they submit.

 

Students are responsible for evaluating the validity and reliability of the resources they use in their academic work, especially when utilizing AI tools. This involves verifying the credibility of sources, checking for accuracy, and ensuring that the information is current and relevant. AI tools may sometimes provide inaccurate or plagiarized citations, and in such cases, any penalties will primarily be the student’s responsibility. Additionally, students must uphold the quality and integrity of their final work, which includes conducting thorough research, providing proper citations, and adhering to academic standards.

 

3. Instructors decide which tools are allowed. Instructors have the authority to decide what resources a student can use on an assignment or assessment. The UNT Health Student Code of Conduct and Civility (UNT Health Policy 7.105) reads that “[u]se of sources beyond those authorized by the instructor in writing papers, preparing reports, solving reports or completing assignments” constitutes cheating. Ideally, these expectations are communicated to students either in assignment instructions and/or in the syllabus. See UNT Health Policy 6.101 “Academic Freedom and Academic Responsibility” for more details.

 

Keep in mind, many Generative AI tools often exist inside comprehensive software packages (e.g. Grammarly, Microsoft Office, etc.). Note that if an instructor has indicated AI is not to be used, you may need to adjust your use of these software packages such that you do not utilize the generative components.

 

If you have doubts or questions about what resources are appropriate for the completion of an assignment or assessment, please review your syllabus or assignment instructions or contact your instructor.

 

4. Use of AI should be open and cited. It is important for students to list the sources used in their work, as uncited resources can be easily confused for the student’s own work. If language or ideas come from AI, that information should be cited. Be mindful that your instructor may also require other documentation or stipulations for disclosing AI use.

 

5. Understand the content. If you use AI to generate content that you do not understand, you may not be achieving the learning objectives of the class.

 

6. Prepare for your career. With the growing capabilities of AI, many professions—including health professions—are seeing AI tools developed specifically for work within those professions. It is appropriate during your studies to prepare to use the kinds of AI tools that will be expected of you as you join the workforce. It is appropriate to rely on your faculty for guidance regarding the use of these tools, and when their use is not appropriate as you engage in learning designed to prepare you for a career.

Artificial Intelligence refers to machine-based systems that, when given human-defined goals, can make predictions, recommendations, or decisions that affect real or virtual outcomes.

At UNT Health, AI is a supportive tool, not a replacement for human expertise.

• In research, AI helps identify patterns in large datasets, accelerates drug discovery, and enhances the reproducibility of results.
• In clinical care, AI aids early disease detection, tailors treatment to the individual, and supports preventive care through predictive analytics.

Human oversight, professional accountability, and ethical review always remain essential.

AI tools bring great promise but also new risks. Clear guidance ensures that UNT Health can utilize AI to enhance outcomes while minimizing harm.

• Reputational Risk: Mistrust may develop in the public or sponsors due to misuse.
• Privacy: Uploading data on patients or the research process in an unsecured application could breach HIPAA or other laws.
• Equity: AI models could exacerbate discrimination against vulnerable populations if not properly developed and implemented.
• Accountability: Using AI and presenting outputs derived from AI to misrepresent research outcomes could constitute research misconduct, as well as malpractice.

1. Consider your role: Guidance is available to support various roles at UNT Health, including researchers, clinicians, students, and staff. When researchers also serve as instructors or clinicians, they should review the guidance relevant to those additional responsibilities to ensure complete alignment with institutional expectations.

 

2. Integrity in Research Practices: Researchers must uphold integrity, accuracy, and reproducibility of their research, especially when using AI tools to support literature reviews, analyze data, or draft manuscripts. Outputs from AI should be evaluated for rigor and validity before being implemented.

 

3. The use of AI Tools Should Be Transparent and Properly Cited: If AI is involved in research outputs, such as creating text, images, or statistical data, researchers must also acknowledge this use. Properly cite AI tools to convey accountability and make it possible for others to recognize how AI has impacted the process. Researchers should also check their funding agencies’, journals’, and ethics committees’ standards to which they are held, as there may be specific guidance on referencing AI.

 

4. Ensure Compliance with Funding Agency Requirements: Federal agencies such as the NIH and NSF set expectations for handling, sharing, and securing data. Use AI tools only in ways that comply with these standards, particularly in terms of data reproducibility, transparency, and privacy.

 

5. Understand the Technology: Researchers must comprehend how the AI systems they utilize function, their limitations, and potential biases. Misuse of AI can give rise to erroneous findings or ethical breaches. Seek training or consult with an expert if you aren’t reasonably sure that you are using the AI system properly.

 

6. Prepare for Ethical Challenges: The use of AI may raise legitimate concerns related to fairness, inclusivity, and bias in its outputs. Researchers should consult their institutional compliance office or human subjects’ ethics committee if they are using AI in research involving human subjects.

1. Consider your role: Graduate students often utilize AI both as learners in the classroom and as emerging professionals conducting independent research. When acting as researchers, such as when conducting data analysis for a thesis or writing a dissertation, students are expected to adhere to the same ethical standards as professional staff, with specific attention to the boundaries of academic authorship.

2. Responsible Use in Thesis and Dissertation Writing: AI tools can serve as effective aids for brainstorming, outlining, or refining grammar, but they must not replace the student’s own critical analysis or authorship. Generating substantial portions of a thesis, dissertation, or manuscript using AI without proper attribution undermines the educational purpose of the degree and may constitute academic misconduct or plagiarism.

3. Verification of Sources: AI models frequently hallucinate citations or misinterpret source material. Students are responsible for verifying every primary source referenced by an AI tool. Relying on unverified AI summaries for a literature review can lead to the propagation of errors and a lack of academic rigor.

4. Data Privacy in Research Training: Students must strictly protect the confidentiality of their research subjects and proprietary data. Never upload unpublished research data, sensitive participant information, or unpublished intellectual property into public AI tools, as this voids data privacy protections and may violate IRB protocols or data use agreements.

5. Transparency and Citation: If AI is utilized in the creation of research outputs, whether for drafting text, processing data, or creating figures, this use must be transparently disclosed. Consult the specific citation guidelines for your discipline and the expectations of your advisory committee to ensure you are citing AI tools correctly and acknowledging their role in your work.

1. Consider your role: Guidance at UNT Health supports multiple professional functions. When clinicians also teach or conduct research, they should review and apply the expectations specific to those roles as well.

2. Patient Safety and Care Standards: The use of AI in clinical care must always prioritize safety and adhere to established professional standards and guidelines. Verify AI recommendations before incorporating them into care plans, ensuring they are appropriate and evidence-based.

3. Transparency in AI Use: When clinicians utilize AI tools in clinical decision-making, they should clearly disclose this use to patients. Patients deserve to understand how AI contributed to their diagnosis or treatment plan—document disclosures in patient records to ensure a clear record of AI involvement in decision-making.

4. Compliance with Regulations: All tools involving patient data must go through UNT Health–approved vetting processes to ensure compliance with HIPAA and other federal or state regulations. Use only approved platforms, anonymize data whenever possible, and avoid uploading patient information into unvetted tools.

5. Understand the Tool’s Capabilities and Limitations: Before integrating an AI tool into practice, clinicians must know what it can—and cannot—do. Misuse can lead to errors or inappropriate care. Seek training and refreshers to remain current on emerging tools and their responsible application.

6. Maintain Professional Judgment: AI is a support, not a substitute. Exercise clinical expertise first, using AI as a complementary resource. Clinicians remain responsible for decisions and outcomes.

7. Adopt AI Responsibly for Career Readiness: Healthcare is rapidly evolving, and AI will play an increasingly significant role in daily practice. Stay informed about developments in AI applications relevant to your field, both to improve patient care and to remain prepared for the future of medicine.