Current Students

The originality of your research from human thinking should always be the top priority. Appropriate use of artificial intelligence (AI) in Earth Sciences research (e.g., data analysis leveraging a neural network) will fit naturally into a Methods section of a research paper or thesis. AI should not replace the act of originating ideas, organizing thoughts, or learning to do scientific reasoning.

The Department of Earth Sciences affirms that graduate research (theses, dissertations, and capstones, for MA, MS and PhD degrees in Geological Sciences and for MS in Computational Earth Sciences) must represent independent, original scholarly work that advances knowledge in the geosciences. Artificial intelligence (AI) tools may be used as research tools, but not as substitutes for intellectual contribution.

All AI use must comply with the University at Buffalo Academic Integrity Policy and reflect norms of the Earth Sciences profession, including transparency, reproducibility, and ethical data practices. Committee Oversight

• The student’s thesis committee must approve any substantive use of AI in research workflows.
• Thesis committees may set additional, project-specific limits.
• Thesis committees are responsible for ensuring that the student demonstrates that the research output makes an independent intellectual contribution.

Pending a discussion with the student’s committee, AI tools may be used in ways that support, but do not replace, scholarly work. Permissible uses include the following:

• Exploring potential methodological approaches*
• Identifying broad topic areas or keywords for literature searches * 

• Assisting with literature discovery* 
• Organizing bibliographies or notes* Writing and Communication
• Editing the student’s original writing for grammar and spelling*
• Formatting references* Data and Technical Assistance
• Assisting with coding, such as debugging student-authored scripts* (e.g., MATLAB, Python, R) 
• Using deep learning in scientific research and computational modeling* (e.g., machine learning, neural networks, surrogate models, model emulators)

*The student must verify all AI-generated material or ideas independently and present such verification to the committee.

Pending a discussion with the student’s committee, the following uses are not permitted because they undermine originality and academic integrity:

• AI as the source of hypotheses, research questions, or conceptual framing
• Generating text, figures, or interpretations that are presented as the student’s original work 
• Original writing, including writing any portion of a thesis/dissertation 
• Generating literature reviews without critical evaluation and verification 
• Producing data, results, or analyses without full methodological understanding (excepting specific neural networks, surrogate models, and model emulators that are well documented by the scientific community)
• Fabricating citations, sources, or datasets 
• Using AI in ways that obscure authorship or intellectual contribution 

This is not an exhaustive list of unacceptable uses.

Students are responsible for understanding how AI platforms store and use input data.

Students must not upload to AI tools:

• Unpublished data (e.g., field data, lab results, model outputs) 
• Sensitive or proprietary datasets 
• Collaborator-shared materials without permission 

All graduate work must include a transparent disclosure of AI use, typically in a methods appendix or acknowledgments section (see appendix item for example):

• Tools used (e.g., ChatGPT, Copilot, etc.) 
• Purpose of use (editing, coding assistance, figure drafting, etc.) 
• Description of how AI outputs were validated by the student
• Representative prompts (if substantive use occurred) 

Students must also adhere to:

• Journal policies regarding AI authorship and disclosure 
• Funding agency requirements (e.g., NSF data integrity expectations) 
• Norms within Earth Sciences regarding reproducibility and transparency 

The department recognizes that AI literacy is an emerging professional skill. This policy encourages responsible, transparent and critical use of AI tools to prepare graduates for careers in academia, government, and industry.

AI is a tool, not an author, analyst, or scientist. All conclusions, interpretations, and scientific contributions must originate from the student. The student is held responsible for the work they produce. Students should review the output of AI and ensure the legitimacy.