Introduction to Ethical Guidelines Section

We also noted a tension between students' strive for speed and efficiency, and the faculty's emphasis on comprehensive and reliable learning processes. We believe that the best outcomes will emerge at the intersection of these two, and that clear institutional guidelines concerning the use of Generative AI will prove handy in both balancing the perspectives of students and educators, and guiding students who are more conservative about the use of AI in Education. 

‍

Ethical Guidelines

Embrace the use of AI in Education as means of preparing for the future BUT bear in mind:

1. Accountability - Students are responsible for critically evaluating AI-generated content. Simultaneously, AI tools should not be used as a sole source of student work, and careful consideration should be put into differentiating a good prompt from a bad one.
2. Unpredictability: Students must be prepared for AI to provide varied responses, even based on the same prompt. Hence, students should treat the outputs of LLMs with a grain of salt, understanding that it might not always respond as expected.
3. Refinability - Students should refine the prompts given to AI tools, particularly if they deliver inaccurate or non-relevant outputs. Furthermore, students should assume that the LLM does not keep track of their prior work, as the context windows are often limited.
4. Confirmation Bias Awareness - Students should be aware that the phrasing of a prompt can influence an AI tool's responses. It is advisable to formulate questions neutrally and to consider exploring perspectives from all sides to ensure balanced and unbiased information.
5. Caution - Students should remain cautious about sharing sensitive or private information with AI, treating their interactions as if they are public. When dealing with academic work, students are advised to turn off the Chat History & Training feature of GPT-4.
6. Collaborative Experimentation - Students are encouraged to openly experiment with various AI models and/or tools to find ones that best meet their needs. Furthermore, students are advised to share their perspectives and applications with peers to augment the collective Generative AI intelligence. ‍
7. Transparency: Students are required to be transparent about the use of AI in academic work. The role of AI tools needs to be acknowledged and specified, not only to maintain academic integrity, but also to improve the educational experience and design for both individual students, and Minerva University as a whole.

‍

Concerns for Equity

As found by UNESCO, just 7% of K-12 schools reported having “institutional policies and/or formal guidance concerning the use of generative AI applications.” And the largest school districts in the US are taking wildly different approaches – for example, the Los Angeles Unified School District has blocked ChatGPT, while in Chicago and New York, public schools are not blocking access to ChatGPT – which not only leads to different AI integration standards, potentially harmful for future research and academic collaborations, but also increases the disparity of education outcomes, making some students far more “future-ready” than others.

‍

The AI Citing Framework

Properly citing ChatGPT and other AI tools is essential for learning, not only helping to uphold academic integrity but also reflecting the ability to critically engage with external materials.

A suggested way to accomplish this is by dividing the assignment into stage and tasks. For each segment, clearly identify which tools or resources you used and for what purpose. This framework not only streamlines the process of both completing the assignment and adequately citing AI tools, but also helps educators understand the extent of AI's involvement in your work.

For instance, if ChatGPT provided initial ideas for a paper, students are required to mention this explicitly. This framework fosters transparency, awareness, and efficiency of the students.

‍