– Welcome. I am Jim Lang. I’m Professor of the Practice at the University of Notre Dame. And I’m here with some folks who have developed an AI acceptable use framework, the Medical University of South Carolina. They’re gonna share some context about that framework, how it’s being used, how it developed, and some examples of how it can be put in practice for educators at another institutions. So I’m gonna invite everyone to sort of quickly share who they are and their role at the university.

– Hi, everyone. My name is Julaine Fowlin and I’m the Executive Director for the Center for the Advancement of Teaching and Learning. And I’m also an Assistant Professor in Academic Affairs.

– My name is Tom Smith. I am the Director of the Center for Academic Excellence and Writing Center at MUSC, which is a one-stop academic support shop for students. And I am also a professor in our Academic Affairs faculty.

– Hi, I’m E’lise Nissen and I’m the Director of AI in Education and Scholarship in the Center for the Advancement of Teaching and Learning here at the Medical University of South Carolina. I’m also an instructor in Academic Affairs.

– All right, welcome. I’m looking forward to our conversation here. Julaine, I’m gonna start with you. So could you tell me how the AI Acceptable Use framework came around about? And so how does it fit within the AI and the ecosystem at your university?

– Thank you so much. First, I wanna acknowledge that this was a beautiful collaborative effort. Tom, myself, and E’lise were just the voices today, but the framework involved several cross-disciplinary and department faculty and staff who contributed to it. At MUSC, AI has never been just like an add-on. It’s part of our strategic initiatives, and we have two priorities. One is to really use AI in a transformative way for healthcare, research, and innovation, and the next is to prepare the future workforce. And so, when AI started, and the buzz was, they were like, “How can we address this issue in a way that faculty and students can have shared language?” So, it started off with Tom and myself, and our former Associate Provost, Gigi Smith. And so, we did the plagiarism, we adapted our plagiarism statement to include some AI. So then when we presented it to faculty, they were like, “We need more. We need a shared language out on how to operationalize this. We know you’re not telling us what to do.” And so, when we look at AI in education, we’re thinking, okay then, you have literacy, you have two access, you have governance and policy. You have curriculum integration, you have these experimentation. How can we empower our faculty and our students to determine how and when AI should be used in a way that is very clear and transparent? And so we got together, we looked at what was out there, and there was ‘The AI Assessment Scale’ that was validated and developed, and then we adopted that for a context. So it’s more about, this is our big system, how can we provide the tools, so faculty and students have that shared framework, shared language, as we move how we’re navigating AI the MUSC way?

– So how was it presented to the faculty? Did it come out from administrative offices? Was it, you know, something that came, sort of bubbled up and was shared? How was it shared with faculty?

– So the first step that we did is that we had to go through a review process. So here at MUSC, we have an education advisory council, which is made up of all the associate deans, or academic deans, in the various departments. So we had to get their approval first. So, when we presented it the first time to them, they actually gave us really good feedback. They’re like, “What if we had an image and how could we get this in the syllabus?” And so, from their recommendation, we worked on a generic communication for faculty. Somebody from my office, Julia Liebenrood, very artistic, created the visual that we will share as well. We were able to put it in Brightspace. And so, the essence of it, as you are in educational development, is how can we give faculty the tools that they need without any work? So everything was literally a copy and paste. We did several workshops, and then the framework also includes some examples. And the first thing that we did is, two people from my office, Mary Smith and Alex Walters, they were teaching a graduate studies course on teaching and they modeled how it could be used. So lots and lots, and lots of support in terms of that to make sure that faculty have the tools that they need and don’t have to think too much, because they have other things to do. So, that’s kind of how we’re thinking about it.

– Yeah, it’s good to see. So, it was a sort of combination of faculty, you know, input and administrative offices, and everyone kind of coming together on this framework. So that always signals something good happening in higher education. So you sort of mentioned this framework comes from this AI Assessment Scale. And actually, when you look at the framework, the document itself has this list of things that have been evolved over the course of the document. So, Tom, maybe you could tell us a little bit from us what those kinds of developments or evolutions is undertaken, from initial thoughts about it to where it is now.

– Yeah, well, and I think this is a little connected to the question about faculty, understanding that it came from not just top-down. It’s understanding the position that faculty are in sometimes, and giving them the freedom to make some decisions about an emerging sort of technology that some regarded as a threat and some would regard as a wonderful new sort of tool for students and for them to use. So, we wanted to remove any connotation from what we were offering that suggested there was a value judgment associated with no AI use versus full-on AI exploration. And so, switching that language from scale, just to be sure we weren’t saying this is like lower order or higher order to framework, like this is how you might frame communicating with students, was part of that shift we wanted to make. And, then, I think another important element of that shift was, at our institution, which is a health sciences education institution, primarily, and because of national certifying exams is really got a view of the word assessment that suggests that assessment is about multiple choice test questions. We wanted to shift the word from this is a framework for presenting AI use for assessments, to thinking of it more in terms of tasks that you might ask students to use in the process of their learning. And so that would hopefully free people up to think of this as comprehensive about the entire learning experience of a course, not just for a graded multiple choice test.

– Yeah, actually, you’re sort of seeing that framework, for example, the AI planning is one of the things you have in there. The AI planning is an example of, like, not just preparing for a test or taking a test, right?

– Yeah, yeah. And I think that, look, anybody who’s listening to this who may be in an academic support center that’s primarily student-facing, knows that a lot of what’s happening in an academic coach or counselor such as myself working with a student, is trying to help the student interpret the language a faculty member has used to describe an assignment or describe a task they have to do. And so, what was on our mind is developing a framework that could encourage that communication and allow a faculty member to say, “Oh, when I say don’t use AI, I don’t mean you can’t use it to brainstorm or plan something, I simply mean you can’t use it to produce the draft you turn into me.” And, we just know that there tends to be imprecise language that’s used to sort of give assignments sometimes, and we wanted to provide a framework that allowed things to be communicated with greater precision.

– If I may add, something as well that became very important with academic tasks, as the team was made up of like instructional designers, faculty, people were supporting. But we realized that, sometimes, within one task, like one assignment or one activity, the instructor may have different AI acceptable use for different parts of it. Because we work with somebody in the College of Dental Medicine, and she said, “I’m okay with them brainstorming for their late review and doing different things with AI, but when it comes to the methodology, because this is the first time they’re learning it, which is one of our principles, skills first,” She wanted absolutely no AI and she wanted them to learn it on their own. So, this framework gave her the ability to say, for part one, you’re using it for, probably, AI planning, but for part two, it’s no AI. And so, the framework does present that opportunity to be really agile and meet the diverse needs of all our colleges.

– Yeah, that’s great. Let’s talk about, you know, sort of concrete practical stuff. So, E’lise, maybe you can sort of tell us what the framework might look like in an instructor’s course or wherever it might appear, and what examples maybe we’ve seen or you’ve seen instructors using it?

– Great. Yes. So, when this work began, I was an instructional designer in our College of Health Professions, and was seeing the need for this, and was so excited that we were able to develop it and deploy it, because it did provide instructors with the words, right? So, the framework gives people the categories and the words to precisely, like Tom was saying, say what they mean when they say no AI or AI planning. And so, that was really exciting. It also opened the door for an instructional designer and an educator to come to us and say, “I don’t know what’s wrong, but I know they’re using AI, and I don’t know what to do about it.” And so it would help give us the language that we needed to unpack, what are you trying to assess in this assessment? What has changed given AI and how can we reimagine what it looks like? And, using this framework to guide, you know, okay, for this part of the assignment, do we think they shouldn’t use AI at all because you want them to develop those foundational skills? Great. No AI. And then move on to the next part of the assignment. So it’s been really instrumental.

And then, personally, I’ve been working with Dr. Greseth in the College of Graduate Studies on a course that we redeveloped. The course is called Scientific Writing is Persuasion. And he came to me and said, “Last time I taught this, it felt like they were using AI,” right? And I could say, blanket statement, for this course, it’s a writing course, we’re not going to use AI, but is that realistic? Is that how people work in? Is that how the real world works? So, we had two goals, right? We wanted to help the students learn how to use AI in ways that made sense in their scientific writing. And we wanted to help them preserve their voice and we wanted to help them not become dependent. Those were the two concerns that they expressed. I don’t wanna lose my voice as a scientific writer and as a researcher, and I don’t want to become overdependent on using AI. I wanna actually learn how to do this myself. And so, the framework was really instrumental in helping us work through what the assignment parts were. So right now, they do no AI in the beginning. So they draft, for example, a description of their research for a mixed audience. And at the beginning, they do no AI, and that gives them a baseline of like, this is what I wrote originally.

And then, every time that we meet, we discuss a different technique. Today was, actually, plain language, like how do you write in plain language for a mixed audience? And they knew that, when they came to class, they were going to have a peer review their writing, and then we were going to use a safe AI prompt to have AI review their writing and provide feedback. And they’re comparing, it’s modeled after the Pair Framework. I know that you’ve talked about this in other “OneHE” sessions, but we’ve modeled it after that and used our AI acceptable use framework to say exactly where in the writing process they should and should not use AI, and how. And the framework provides that nuance that’s needed. So, we have AI limited where they are using a prompt that we provide to get feedback on their writing, and it’s been really eyeopening for them. And so, if we didn’t have the framework, it would be more difficult to say, over the next five weeks when we’re doing this work, how and when can you use AI, and in what way? And, we also think that it’s providing psychological safety. And we’re actually studying this – if psychological safety is increased as a result of the transparency that the framework allows. So, lots of interesting good work going on.

– Just tell me a little bit about that. You mean, in terms of like they’re anxious about being accused of AI-

– Yeah.

– Or like… Yeah, is that what you mean?

– Absolutely. So they’re worried, right? Like, maybe I don’t use it and I’m accused of using it. Maybe I use it in a way that I think is appropriate, but then it wasn’t. So, one of the greatest things that have come out of the framework is that students have actually come up to me after class and thanked me. Thank you for discussing this openly. Thank you for being clear and transparent. Thank you for talking about what AI can and cannot do in terms of writing. So, it’s opened the doors to all of those conversations, and it’s created an environment where people feel like they can talk openly and safely about their AI use.

– I wanna just finish here, maybe thinking about going beyond the classroom to a framework like this maybe applying to ourselves or, to like writing for research journals or something like that. Like, have you thought about the possible ways that it could have impact on faculty work or like, you know, writing for publication?

– I do think if journals adopted it, ’cause several journals that I’ve looked at, they’re very generic in how they use it. And, I was even working with a research faculty and she asked me to review something, and I said I can review it, but then I could ask AI to help me, like AI undertenant. So like, at MUSC, we have Copilot to review. And she’s like, “No, no, no, no, no.” So faculty, on their side, they’re experiencing fair. And I don’t think the journals have come to a conclusion. At MUSC, we were very clear that we do not support the use of AI detectors. If a faculty wants to use it, that cannot be their only source of truth. But, some journals are using AI detectors. And so, there is not a shared mindset or framework in that space that I do think adopting something like this. Like, if APA decided, “Okay then, let’s adopt this. Just like how we have citation guidelines, can we have some standardized AI guidelines, because it would really help. So that we don’t feel like when we’re putting in a paper, we’re cheating.” I always put how I use AI in an article. So, like the article, about empowering educators, we used AI and we were very transparent about how we used it. So, that’s a good question, and something that maybe you can start at project, Jim? And we can collaborate.

– Well. And, one of the things I’ll point out about the scale, is that we’ve got a feature that I like, which is a documentation section in the scale, which, basically, suggests to faculty and their communications with students, that disclosure and reflection about use of AI is really an important element of this process, particularly at this stage of AI’s emergence into people’s work lives and learning lives. And, with conflict of interest statements, that we are asked to complete as faculty, I think we’re used to the idea of disclosure as being sort of one of the ways that we put things in the sunlight in order to disinfect. It might be worthwhile for us to provide a space and a framework within journals, within conferences, to sort of actively disclose any use. And if that is, in fact, a prohibited use, then, of course, the journal can say, “Well, thanks for disclosing. Please send to some other journal or some other conference.” But, I think because of the uncertainty around citation and what exactly that means in the context of AI, disclosure is really a critical element to feel ethical about the intellectual work that we are doing.

– Yeah, that sense of psychological safety. Safety is also important for faculty members who are now also concerned that they’ll get accused. Because there have been some very public cases of this recently, of authors, both public writing and in academic writing. So yeah, a framework like this could be helpful for that. Okay, so that’s our time. Really grateful for you all. Not only creating this framework, but offering it up to the public to sort of adapt it and use it in their university as well. So, the framework should be available under this video or like as a link in the video. So we sort of encourage people to take a look at it and consider whether it might fit in your institution. Thanks to all for joining me today.