AI Task Force for Teaching & Research

The task force’s key policy recommendations across Boston University are as follows:

• Critical Embrace: BU should not universally prohibit or restrict the use of GenAI tools. Rather, BU should critically embrace the use of GenAI, support AI literacy among faculty and students, supply resources needed to maximize GenAI benefits in research and education, and exercise leadership in helping faculty and students craft adaptive responses.
• Pedagogical Clarity: GenAI policies adopted by any college, school, or departmental unit should be consistent with the University’s policies and reflect the critical embrace of GenAI technology. Consistent with academic freedom, individual instructors should be free to define GenAI policies suited to the learning goals of their courses, and every syllabus needs to state the instructor’s GenAI policy. Consistent with citation practices, instructors and students should acknowledge use of GenAI tools.
• Academic Misconduct: BU should advise instructors to exercise caution when using GenAI detectors as evidence of GenAI use. GenAI detector output should be regarded as only one part of a wider evidence base in evaluating possible academic misconduct. If used, GenAI detectors should be applied equally and fairly, and faculty should be aware of selection bias when applying GenAI detectors to specific suspected cases. Advance notice should be given in syllabi, including naming the specific detectors employed, so that students have an opportunity to use them also. Instructors need to be informed and supported in using reliable and consistent detectors.
• Security and Privacy: BU should adopt policies to prevent the inadvertent publicizing of sensitive or valuable information through uploading it to GenAI tools.
• Centralized Decisions: BU should centralize decisions on GenAI tool acquisition and licensing, on resourcing and personnel for supporting GenAI literacy and pedagogical reflection, and on administrative structures to ensure ongoing adaptive response to rapidly developing GenAI technology.

GenAI has become commonplace in creating artifacts, including but not limited to areas that incorporate an iterative invention and revision process such as writing, designing, and creating art. The BU AI Task Force recommends BU should support GenAI not as a replacement for the process of creation, but rather as a collaborative tool. GenAI tools can be effective in these areas if integrated with human oversight, and if students are taught how to use them appropriately and responsibly.

Key Takeaways

• GenAI Requires a Nuanced Understanding: Faculty can guide students in understanding how GenAI tools can serve as a part of their creative toolkit in courses that incorporate writing or designing, emphasizing its supplementary role rather than a substitute for human creativity while also demonstrating GenAI’s limitations. This will foster an AI literacy that appreciates GenAI tools only as a potential partner or assistant in writing and other creative activities.
• Teaching Courses with GenAI: The application of GenAI tools in the teaching of courses that incorporate creative processes extends beyond mere content generation. It offers an opportunity for personalized learning experiences, exposing students to a wide array of styles and techniques. For instance, in tasks that incorporate writing, GenAI tools can assist with language development, genre awareness, and the structuring of arguments.
• Design of Process-Oriented Assignments: Designing assignments that require an iterative creative process that integrates GenAI demands particular care and consideration. This involves creating assignments that leverage GenAI as an educational tool while also fostering critical thinking and creativity in students. For example, faculty may design writing assignments where GenAI tools are leveraged only for initial idea generation and brainstorming. Such assignments would then require students to select one concept and develop and refine it further. Other assignments may ask students to reflect on their process of working with GenAI tools.
• GenAI as a Collaboration Tool: The existence of GenAI tools should be acknowledged in course policies. It is important that the emphasis in courses that incorporate writing and design elements should be on the process rather than solely the end product. The necessity of maintaining ethical standards and human oversight will always remain, and GenAI should be framed as a collaboration tool and assistant.
• Use Cases: GenAI has been successful in assisting with academic writing, visual arts, tailoring learning materials to individual student needs, graphic design, integrated communication campaigns, and public-facing campaigns.

The promise of GenAI tools in research is considerable, as are the challenges. It is important to enable BU researchers to use GenAI to advance their research and to address major societal challenges, while also mitigating any problems that may result.

In addition to researching the landscape of GenAI technologies, referencing expert opinions, and analyzing leading academic recommendations, the BU AI Task Force surveyed 15 associate deans of research (nine provided responses) to further tailor recommendations to the needs and concerns of the Boston University community.

This research guided the Task Force’s benefits and concerns about AI in research, as listed below.

Key Takeaways:

• GenAI Can Help Save Time: GenAI in research has the potential to save researchers time by synthesizing, analyzing, and automating information, including with the ability to extract information from text, summarize passages, automate data collection, and identify relevant publications for further research.
• Technical Support and Coding: GenAI tools can automate technical-support activities by answering technical questions and providing instructions on how to perform a plethora of computer-related and other tasks. GenAI tools can also generate computer code in almost any programming language, greatly accelerating and simplifying tedious and time-consuming computer-programming tasks.
• Disciplinary Applications: GenAI tools have started to be used in various scientific fields. Examples include: aiding healthcare and disease diagnosis, robotics, material science and chemistry, and more. GenAI tools can also compose text for scholarly papers/presentations/proposals; take interpretative summary minutes in meetings, when coupled with automatic speech recognition; generate illustrative graphics; and more.
• Responsibility and Accountability: BU should consider strengthening guidelines for responsible conduct of research and implementing a mechanism for ethics review of research projects involving AI, much as BU currently uses an institutional review board to oversee human-subjects research.