Syllabus Guidance (@ UAPC, In Discussion)

The advent of Artificial Intelligence (AI) has sparked a great need for clarity on how its use (or restriction) is managed in Rider’s curriculum. As part of a comprehensive effort to address AI at Rider, members of Academic Affairs leadership are requesting the approval of official language for inclusion in all course syllabi as part of academic policy. This effort was spearheaded by CEHS Dean, Dr. Jeannine Dingus-Eason, who created a survey for full- time faculty suggesting language for three different policies surrounding AI (e.g., permitted, some permission, not permitted). We had 114 faculty respond to this survey as follows: 54% indicate some permission, 29% no permission, and 17% full permission, demonstrating our approach of providing multiple policies best suits faculty needs.

Suggested language:

Permitted

" You may use AI to complete the work in this course. More specifically, you may incorporate ideas, writing, or other content that has been generated by AI in your submissions; however, AI generated content must be properly cited, differentiating your own ideas and writing from the ideas and writing that are AI generated. Note that you are responsible for verifying the accuracy of AI generated content before including it in your submissions."

OR

"This course encourages and embraces the ethical use of Artificial Intelligence (AI). You are expected to actively incorporate AI tools while upholding integrity. You must assess the integrity and impartiality of your submissions, ensuring they remain unbiased. AI has inherent limitations, and human supervision is necessary to verify the quality and appropriateness of the output."

Some Permission

“I see emerging AI tools as helpful for learning, but not a substitute for your own creativity. I will help you explore how to use AI responsibly and ethically within this course”

OR

“AI use is allowed for brainstorming and critique, as specified in assignment prompts. All AI generated material must be cited using the appropriate academic style.” OR “You may use AI for idea generation or research. Final submitted work must be your own and not AI-generated”

Not Permitted

“All generative artificial intelligence tools are strictly prohibited in this class. You are expected to complete all of your work without the use of any generative AI tools. Students turning in work violating this policy will be subject to all academic and disciplinary procedures associated with plagiarism and cheating.”

OR

"In this course, we learn from our experiences and from each other. You are encouraged to play with your creative freedoms, to problem-solve, and to contribute your ideas in your unique ways. Owing to our commitment to authenticity in this course, the use of generative artificial intelligence (AI) to complete coursework is prohibited. All work presented by you for a grade in this course must be of your own doing."

Draft of IRB Guidance Regarding AI and Zoom (@IRB, In Discussion)

The Rider IRB acknowledges that the emergence of AI presents researchers with an opportunity to use AI to aid in collection, creation, and analysis of research data.  Currently, there is not a consensus on best practices regarding human subjects research and the use of AI.  Given the fluidity and the ephemeral nature of AI technology, the IRB offers the following guidance and protection for both the researcher and the participant.

When conducting research, a researcher may  utilize AI tool in human subjects research:

• The AI tool collects data from humans through interaction or intervention.
• The AI tool may be used to help create documents and assist in drafting questions, as well as aid in obtaining informed consent from human participants;
• The AI tool is used to obtain, analyze, or otherwise access potentially identifiable information about participants.

The IRB recognizes that the data that AI tools have access to should be carefully protected. Unless the AI tool has been vetted by the Rider OIT Security, there is no guarantee that the information AI tools access will remain confidential. This is why the Rider IRB recommends the use of a localized AI tool (e.g. Ollama, or GPT4ALL), which protects your participants' privacy, does not require the Internet, and the data does not leave your computer.  Both Ollama and GPT4ALL, which are mentioned, are free to download and use.

More broadly, the Rider IRB continues to recommend collection of personal identifiable information be limited to only what is necessary to complete the research study. 

The recommended guidance provides researchers with consent form disclosure, but through the IRB submission process an additional AI supplement must also be completed. 

Recommended Guidance:

The Rider IRB recommends and in most cases will require that researchers disclose to their research participants the use and context of AI in the consent form documents, if they plan to use AI.  In addition, the Rider IRB recommends and in most cases will require, that researchers disclose the following for the participant to agree that the use of AI is permissible: 

1. Describe in layman's terms how the AI tool works
2. Describe in layman's terms how and why the AI tool will be used in the research study.
3. Describe in layman's terms how the AI tool may interact with the participant’s provided data.
4. Describe risks to participants associated with the use of the AI tool.
5. Describe and provide links to any known privacy and/or data confidentiality policies for the AI tool.

AI Supplemental Form

1. Name of AI tool:
2. Is the AI tool: commercially available with no modification, commercially available with modification, purpose built
3. Is the AI tool connected to a cloud that is accessible to people external to Rider?
4. What part of the methodology or research protocol is the AI tool involved?
5. How do you plan to mitigate any risk or bias that using the AI tool may bring about?
6. What participant data will the AI tool collect, access or analyze?
7. What are your planned reporting mechanisms for communicating issues with the AI tools to study participants?

Zoom Guidance

To safeguard research participants and collected data, at a minimum, the following guidelines should be followed:

• Researchers should recommend that participants protect their privacy by completing research activities in a private and quiet space.
• Researchers should ask for consent prior to using Zoom Cloud for storing data, pressing Record on a Zoom session, taking screen shots, or allowing the Zoom Cloud service to collect transcriptions.
•  Researchers should set up Zoom meetings with password protections.