Artificial intelligence (AI) tools are increasingly used in research, teaching, writing, coding, data analysis, and creative work. These tools can enhance productivity and support learning and scholarship, but they also introduce risks related to accuracy, authorship, academic integrity, privacy, security, and intellectual property. Graduate students are responsible for ensuring that their use of AI complies with University policies, disciplinary norms, applicable requirements for citation and acknowledgment, and the expectations of their faculty advisors, instructors, research groups, and degree committees.

TOPIC INDEX

Before Using AI
Core Principles
Publishing and Presenting Research
Additional Resources

Before Using AI

Consult Your Department

Expectations regarding AI use vary across disciplines. Some departments, programs, research groups, and laboratories have established guidance governing the use of AI in coursework, research, publications, creative work, and dissertations. Graduate students should first consult their department or program to determine whether discipline-specific guidance exists. A template for use by departments can be found here.

Discuss Expectations with Your Advisor and Committee

Students should discuss AI use with their faculty advisor and, when appropriate, members of their thesis or dissertation committee. These discussions should clarify:

  • Whether and how AI may be used in research, writing, coding, image generation, data analysis, or other scholarly activities.
  • What forms of AI assistance are acceptable and which are not.
  • When disclosure or acknowledgment of AI use is required. Any additional discipline-specific expectations.

Students are encouraged to document these expectations and to share a summary with their advisor and committee members so that all parties have a common understanding and written record of the agreed-upon practices.

Core Principles

You Are Responsible for Your Work

AI tools can generate inaccurate, misleading, incomplete, or fabricated content. AI systems may invent references, citations, quotations, data, analyses, or factual claims.

AI systems are tools and cannot be authors or assume responsibility for scholarly work. They cannot be held accountable for the accuracy, integrity, or originality of research or for ethical and scholarly judgments made in conducting it. Graduate students, faculty, and other researchers remain responsible for the work they create, submit, publish, or present, regardless of whether AI tools were used in its development.

Use AI as a Tool, Not a Substitute for Scholarship

AI can be useful for activities such as:

  • Editing and improving writing
  • Grammar and spelling review
  • Brainstorming and exploring ideas
  • Assisting with programming and debugging
  • Generating figures, graphics, or visualizations
  • Finding, organizing, or summarizing information

AI may assist with these and other scholarly activities, but it should not substitute for the critical thinking, disciplinary expertise, scholarly judgment, verification of sources and evidence, data interpretation, or original intellectual contributions expected of graduate-level work.

Protect Sensitive and Restricted Information

Students must protect University and research information when using AI systems. Students should consider the UC Protection Level of information before providing it to an AI system. UCSB classifies Institutional Information and IT Resources from P1 (minimal protection requirements) through P4 (the highest protection requirements). Students should consult the UCSB UC Protection Levels guidance to determine the appropriate classification of information.

Students should use AI platforms that UCSB has approved for the applicable Protection Level. The University maintains a current list of GenAI Tools at UCSB, including the Protection Levels for which each platform is approved. Additional requirements may apply to highly protected information; for example, UCSB currently requires consultation with the Office of Information Security before P4 information is provided to an AI tool.

Publication status alone does not determine whether material may be provided to an AI system. For example, a student may use an appropriate AI tool to assist with their own unpublished writing, provided that the material does not contain information requiring greater protection and that the use is consistent with applicable policies and agreements. Students should not assume, however, that they have authority to provide unpublished, proprietary, or confidential work belonging to collaborators or others to an AI system.

Before using an AI platform, particularly one that has not been approved by UCSB for the intended information, students should review its terms of service and privacy and data-use policies. Students glean from these documents how information submitted to the system may be stored, retained, shared, or used; whether it may be used to train or improve AI models; and what rights the provider claims over submitted content and generated outputs. Terms and policies can change, so they should be reviewed periodically. Documents containing this information are commonly called Terms of Use, Terms of Service, Data-Use Policies, and Privacy Policies; there is typically one or more such documents associated with each service or overarching product that are made available at sign-up or via the vendor's website.

Students must comply with all applicable research agreements, IRB requirements, sponsor requirements, data-use agreements, intellectual property and confidentiality obligations, and University policies. Students should inquire about relevant requirements with their advisor and collaborators and seek guidance when they are uncertain whether particular information or an AI service is appropriate for the intended use.

Do Not Outsource Scholarly Evaluation or Judgment to AI

AI should not substitute for scholarly evaluation, professional judgment, or other responsibilities entrusted to a graduate student. For example, students should not ask an AI system to independently evaluate a manuscript, proposal, thesis, application, student work, or other material that they have been asked to assess unless AI use is permitted in that context.

This concern is distinct from the privacy and security of the information involved. Even when information can appropriately be provided to an AI system, delegating an evaluation to AI may be inconsistent with the student's scholarly or professional responsibilities. Conversely, providing confidential work to an AI system may be inappropriate even when AI is used only to assist, rather than perform, the evaluation.

AI may assist with evaluation when such use is permitted by the relevant instructor, advisor, review process, agreement, or policy, but students must independently assess the output and remain responsible for the resulting judgment.

Be Transparent

Transparency is essential to maintaining research integrity. Students should disclose AI use whenever required by:

  • Journals, conferences, publishers, or funding agencies
  • Departments, instructors, advisors, or committees
  • University policies and guidelines

AI-generated or AI-assisted text, images, code, analyses, or other content that contributes substantially to scholarly work should be appropriately acknowledged or disclosed in accordance with applicable policies, disciplinary norms, and publication requirements.

Publishing and Presenting Research

Many journals, conferences, competitions, and publishers have established policies governing AI use. Before submitting a manuscript, proposal, presentation, poster, competition entry, or other scholarly work, students should review the relevant venue's policies regarding:

  • Permitted uses of AI
  • Disclosure or acknowledgment requirements
  • Authorship requirements
  • Restrictions on AI-generated or AI-assisted content

Failure to comply with applicable publication or submission policies may result in rejection, retraction, disqualification, or other consequences.

Additional Resources

Because AI technologies and policies are evolving rapidly, students should remain informed about changes in University policies, disciplinary expectations, and publication requirements. The following resources provide additional guidance on the responsible use of AI in graduate education and research.