Category: higher education

Show Notes

• Kauffman, K. (2025). Navigating Choppy Waters: Key Legal Issues College Faculty Need to Know. Bloomsbury Publishing USA.
• Luzianne Iced Tea
• Banana Lady
• Copyright and Artificial Intelligence – U.S. Copyright Office
• Goldstein, Evan (2026). “The Ivy League Cheating Scandal No one Wants to Talk About.” The Chronicle of Higher Ed. an article about the Brown economics professor who experienced extensive student use of AI on a midterm exam in Spring 2026.

Transcript

Rebecca: Faculty create large quantities of content, but aren’t always aware of their intellectual property rights, or lack thereof. In this episode, we explore legal issues associated with copyright law.

[MUSIC]

John: Thanks for joining us for Tea for Teaching, an informal discussion of innovative and effective practices in teaching and learning.

Rebecca: This podcast series is hosted by John Kane, an economist…

John: …and Rebecca Mushtare, a graphic designer…

Rebecca: …and features guests doing important research and advocacy work to make higher education more inclusive and supportive of all learners.

[MUSIC]

John: Our guest today is Kent Kauffman, a Professor of Business Law and MBA Programs Faculty Liaison in the Doermer School of Business at Purdue University Fort Wayne. He is also the author of Navigating Choppy Waters: Key Legal Issues College Faculty Need to Know. Welcome back, Kent.

Kent: Thank you very much for having me. I’m always glad to talk with the two of you, and I appreciate the invitation again to be on your awesome podcast.

Rebecca: Today’s teas are:… Kent, are you drinking tea with us today?

Kent: I am drinking tea with you today. Due to the heat here in Indiana, I don’t know if I’m pronouncing it correctly, but I’ve been drinking the Luzianne iced tea, where the tea bags are turned into an iced tea. So that’s my tea of choice as of late.

Rebecca: Something nice and cool seems like a good option in the heat.

Kent: Absolutely.

John: It’s a hot summer pretty much everywhere. And I am drinking a blueberry green tea.

Rebecca: Green must be the theme in here because I also have green tea today, John. I have a jasmine green.

John: Very good.

Rebecca: Looking for something lighter for the summer. [LAUGHTER]

John: So we’ve invited you here today to discuss intellectual property issues for faculty. One issue that comes up fairly often for faculty is the question of ownership of course materials that are created by faculty. Who owns the copyright to course materials that is stored in a learning management system.

Kent: Well, I think that’s a really needful and urgent question, and to me, it sort of parallels the urgency of academic freedom issues that faculty face right now. Because both of those issues involve finding out well, what what are my rights, whether it’s my right to teach and research the way my discipline respects, or my rights with respect to the work products I create that are attendant to, in this case, my teaching and also my scholarship as well. So if I can just do a quick tutorial, and I’ll try not to take too much time on this. I think of intellectual property like three siblings and a cousin that’s like invited to Thanksgiving dinner, but isn’t really a part of the family, and the siblings are trademarks and patents and copyrights, and that distant cousin is the trade secret. And the difference I see is this: intellectual property traditionally gets its value from being publicly displayed. That’s the point. What is the value of a song if no one’s ever heard you play it or sing it? What’s the value of a script if no one’s ever said it somewhere? And so intellectual property law, federal law, by statute, protects those particular aspects of our creative property. Trade secret law actually is protected by you, and so there’s a completely different way that we get value out of hiding something that we think is valuable, and that’s trade secret law. For us and copyright, for which faculty more than any other intellectual property, notwithstanding some faculty might deal in patent creation, but that’s a whole ‘nother discussion. But as it concerns copyrights, I think the first question to ask is, “What are we creating that is actually worthy of copyright?” And that’s not to be insulting. That’s just to say, by way of background, copyright does not protect our ideas. Don’t tell someone your great idea. Keep it to yourself. Copyright law protects the expression of your ideas. What the statute calls when your ideas are “put into a tangible medium of expression,” or what I might say, a fixed medium of expression. Your podcast is an example. By talking into this recording device, there’s copyright created automatically because this is a recording of the expression of our ideas. So once we understand that copyright protects our work automatically, we don’t need to seek a copyright registration to get that protection. The beautiful thing about it is it protects it automatically once it’s put into a fixed or tangible medium of expression, but the next question is, and this is the perhaps insulting part, is “What are we doing that’s worthy of copyright protection?” Because that requires creativity, and there is a Supreme Court case involving the phone book to which many younger listeners might say, “What’s that?” [LAUGHTER] But there was a lawsuit over the phone book because a competing company was using the white pages out of a phone book, and that went to the Supreme Court of the United States, and they had to answer the question, “What here is worthy of copyright?” And just to cut to the chase, the yellow pages are protected by copyright, but the white pages are not because there’s no creativity in assembling people’s names and addresses. And one feature the court mentioned is that copyright requires that your work have a spark, that was their word or a modicum, their word, of creativity. So if we think about that, the question then becomes: “Am I doing something that is put into a tangible medium of expression that is worthy of copyright that is instructional materials?” The foundational document probably is the syllabus. I think of the syllabus as the starter yeast of academia because it began somewhere else and it keeps getting handed over and over again as new things are added to it. And so, where I work, my syllabi is longer than it’s ever been – not by design, but because new parts must be added to it. Well, I don’t own a copyright in the works of others, so we jump out of that for a second. If I’m creating PowerPoints or tests or any other materials that come from the publisher of the textbook I might use, then I have no copyright in those materials whatsoever. So then we say, okay, is my syllabus copyrightable if I have enough of my own original content in it. That’s a worthy question.

If you think that you have enough originality, creativity, that your compilation of facts, which can qualify for copyright, are original enough, then you might be able to say it belongs to me. There can be a problem with that because one might argue that the syllabus is the department’s document as opposed to your document, since even if you’ve created the class, it’s still under the aegis of your department and your university as well. But in terms of these instructional materials, whether it’s a syllabus or PowerPoints, tests, really interesting essay assignments or projects that students do, I would argue that a lot of those can be worthy of copyright protection because they are first put in a fixed medium of expression because they’re written down digitally or on paper, and I would say that really, for faculty that care about what they’re doing and trying to make a vibrant life out of their classes, absolutely they can be protected by copyright because they meet those modicum standards, those spark standards of creativity. So yes, I would say this, from an initial standpoint, if you believe that your work meets that expectation of creativity, then you could make the argument that your instructional materials could be copyrightable and belong to you, but there’s this however, and the however might lead to the next topic, which is instructional or institutional policy.

Rebecca: I can imagine that if folks were looking at copyright law at face value, they might assume that this would fall under works for hire because it’s part of your job, maybe, to create instructional materials. So, can you talk a little bit more about where this role of institutional policy kind of dovetails with the law here in determining who has the copyright to these materials?

Kent: Yeah, there’s two issues there to think about, and one is policy, and the other is works for hire. And if I can just swing around to the policy issue, and then get to works for hire next. The beauty from an employment standpoint is that universities, and I’m saying this not as an employer, but from their standpoint, the beauty is that they can create institutional policy that also covers intellectual property. And one way that a lot of universities cover our instructional materials is that they will define categories of work we create at a university, and those are usually three types: traditionally administrative, and then scholarly, and then instructional. Now, someone like me who’s really a big believer of scholarly teaching… One could argue that teaching materials or instructional materials could fall into the category of being scholarly, but that’s again another discussion. As it concerns administrative materials. If you at your university are doing a teaching evaluation of a tenure-track faculty member, no matter how well you write that letter, it’s likely not to belong to you because universities categorize those documents as simply being university-owned property. As it concerns scholarly works, peer-reviewed papers, poster presentations, books we write. What I find interesting there is that: 1. universities tend to exempt themselves from ownership of those to start with, which I think is almost a moot point because I have to transfer copyright to those works anyways to the publisher generally to get them published. And there is open-access publications, I realize that, but still we live in a world where lots of our publications are conditioned on our releasing our copyright. So my university can’t claim ownership to my books, for instance, because I don’t have any ownership [LAUGHTER] to my books as well. So then the next thing is the instructional materials, and here we get a situation where universities do something that I think I’m not comfortable with it. But what happens is they often will say you own your instructional materials, but you have granted to us, by condition of employment, a non-exclusive license to use your materials, and the key there is, if you look at your institutional policy, if it has that, is it will often say non-exclusive because an exclusive license, similar to a transfer of copyright, requires both signatures. I must sign, and so must the proper agent of the institution sign as well. But a non-exclusive license doesn’t need signatures. And so, for instance, Purdue’s policy, the university-wide policy, not just where I work at Fort Wayne, but across all of Purdue institutions, is a non-exclusive, royalty-free, perpetual license. To me, that means that as it concerns my instructional materials that are in… and Purdue uses what’s called Brightspace as its LMS… all of that is, in my view, effectively co-owned by Purdue, not owned because they granted me the right to keep my copyright, which is an interesting language phenomena to say we allow you to keep something that you already own as the author of the copyrighted material, but a non-exclusive, perpetual, royalty-free license does mean that they can use my materials, everything in Brightspace or anyone’s LMS for as long as they want, no matter if I work there or not. Some institutions limit that, they might provide a royalty structure if they’re using your materials in a class you don’t teach. Other institutions might say we will cease using your materials if and when you cease working for us. And those are nice, gracious ways of sort of managing that. But to faculty that put their instructional materials on an LMS, as almost all of us do nowadays, I would say yes, you still would own what is worthy of being owned for copyright purposes, but you sort of now have a co-owner, effectively by your employer, who has taken a non-exclusive or maybe even an exclusive license in them. But the work for hire situation is a different angle to this. There’s two ways to create a work for hire, which would mean that the author is not even the copyright holder. They’ve transferred the copyright altogether. The first is the easy one, and that’s where I transfer my work as an independent contractor to the hiring party, and that requires two signatures there.

A couple years ago, I wrote scripts for some educational cartoons for a business law course that was online, and I created the subjects for what they were looking for. They approved them. Then I gave them ideas, and they approved them. Then I wrote scripts for educational cartoons, and they approved those as well. Well, none of those I own because that contract, being a work-for-hire independent contract, meant that they own all my work. The other way is what is known as what creative works that we produce in the scope or in the course of our employment, and that’s another part of the copyright code. And the scope of employment doctrine is an agency law doctrine that says that if I’m hired to do certain types of things that might involve creating intellectual property, my employer gets them automatically, not by needing a signature because I’m creating them in the scope or course of my employment. For faculty, an outsider might say, “Well, everything you do is in the scope of employment, and therefore, wouldn’t your university own them all? Well, yes and no. And to the no part, I’m not a greeting card writer at Hallmark, so I can’t disclaim some of my really pithy statements and say, “Well, I own this one because I’m hired to do that.” But as agency law looks at what faculty do, as it concerns instructional materials, it’s a confusing world. Is that in the scope of my employment? There’s no consensus on that. Some courts will say it is in the scope of your employment, particularly as it concerns curricular materials. There’s a case, I find it kind of sad, where a woman… she was a department chair at a two-year college tasked with creating a whole new program at her college within her discipline… and eventually she was let go and she brought an age discrimination claim and along the way she filed copyright registrations for the curricular documents that she created. One doesn’t have to do that to own them, but one needs to do that if they’re going to sue someone for infringing their copyright. Well, she lost on both counts and, as concerned the copyright issue, the court said part of her job as a department chair included doing things like creating new curricular materials, and those are clearly the work of her employer. But for regular faculty who are creating courses, not new courses by way of a grant, so if you’re hired to create a new course by way of a grant, that’s a work for hire in the first category. You’re going to sign a document and give up your rights to that material. But for the work we create that’s regular instructional work, there isn’t consensus, in my opinion, legally on whether that really is work for hire, and therefore we don’t own it, anyways. But if there’s an institutional policy that actually carves out the whole work-for-hire concept. So for instance, I know that my instruction materials are not work for hire at Purdue because the institutional intellectual property policy already says I own them and grant them a license. Well, therefore, it’s not work for hire because they said I own them; in a work-for-hire situation, I wouldn’t own them to start with.

John: I know in the SUNY system, any work that’s created for online classes, as part of an agreement with the union, copyright is maintained by the faculty members, however, at least a number of institutions have worked with adjuncts to develop courses as works for hire as a way around that. So that we have a mix of both of those in SUNY. I think that Oswego, we allow the creators to own the copyright unless there’s a work-for-hire arrangement that overrides those conditions. So most of the issues that I’ve seen have been in terms of online content that’s created. But what about lectures that are delivered in person? Do you own the rights to those?

Kent: There’s the question. I enjoy face-to-face classes. I certainly am happy to have online classes, and I take them very seriously. I want to make them vibrant, but I still really enjoy being in a room with people and having a dialog. Here’s an interesting case. So there was a lady who called herself the Banana Lady. You can look her up online. She dresses in a banana costume, and she does like corporate events. But what she really does is she tends to sue people who take pictures of her at her events and claim that these are copyright infringements. And in the federal court system here in Indiana, the Seventh Circuit Court of Appeals not only ruled against her, but one thing that they said was she was trying to sort of gin up litigation. But they said that people taking photos of her aren’t infringing on her copyright because she wasn’t recording herself doing her banana woman sort of dance routine stuff. So that’s the starting point. If you’re not recording yourself in class, then you’re not creating a tangible medium expression of your teaching, no matter how creative it might be. So, A) are you recording it? B) is who’s recording it, because if your university is recording it either automatically or that you use university technology to record it, I would regrettably argue that that’s not your property. It’s probably your employer’s property if you’re using its technology as the conduit for creating a tangible medium of expression. If you’re recording it yourself and your university is recording it, I think you could claim you have ownership over what you yourself are recording. But if you’re not recording it, there’s no copyright, and that goes to everything. If you put notes on a blackboard, you’re putting it in a tangible medium of expression. But if you’re erasing it, it’s gone anyways. And so the question becomes: “Well, then, what do students own if they’re taking notes from what I’m doing in class?” If they’re taking notes from what you’re saying and you’re not recording it, they have a valid claim that they own their notes. Again, something can be not worthy of copyright, but still be your property. Copyright is a type of property, so I could own something that’s not worthy of copyright. That’s what trade secrets are all about. So that a business who has a client list they believe is a trade secret property that’s not copyrightable, but it’s still their property. So I’m not saying a student owns the copyright in their class notes because that’s a copyright creativity question. But they could argue they own those notes if they’re taking notes off of what I’m saying. If they’re taking notes off of my powerpoints, and my powerpoints belong to me. I would argue that they’re actually just simply transcribing what is already mine, and they don’t own that. There are some cases, very few, but they’re out there, where faculty have sued publishing companies because in some really large universities, publishing companies might sell you notes for the classes that you’re taking that are highly common large enrollment classes, and faculty have sued saying that’s actually my property, but students are selling those notes to publishing companies, who then create study guides out of them. That’s a really thorny question because if they’re simply transcribing what is your property, you’re not transferring your property to them. So that would be a copyright infringement. One area is what’s called transformative use, which is a part of what’s called the fair-use doctrine, which allows a person to raise a defense to copyright infringement by arguing a statute in the code called the fair-use doctrine. And fair use is about limited usage of some of those copyrighted work for reasons that not are allowable, but are part of the variables for determining if you’re committing a non-violation of a use of some else’s material. Transformative use says if I use your material in a way that’s completely different than how you used it, I have transformed your purpose into a new purpose.

So you could perhaps find a situation where a very creative student is creating a graphic novel out of their class notes that might qualify as transformative use, and they might say that’s a brand new thing, and it might belong to me separately. So just to sum that up, if you’re not recording your work, then someone could claim ownership in it, and that could include notes. If you’re in a situation where students are taking down as notes what you’ve already created in a tangible form, generally speaking, that’s not their property. That’s your property.

Rebecca: So, on the theme of recording, we’ve talked a bit about posting materials on the learning management system and maybe sharing through the learning management system. But many faculty also use third-party sites like YouTube or Vimeo that have licenses very similar to what you described at Purdue. [LAUGHTER]

Kent: Yes, right.

Rebecca: So, how does copyright work for those situations or ownership in those contexts?

Kent: That’s a really interesting and strategically powerful question, and I show my students when it comes to the section in business law and intellectual property, I show them Instagram and Meta’s terms of service for example because what they may not realize is that they’re actually granting co-ownership of their property to those sites, and the language is quite broad. It’s hard to find, but it’s quite broad. So you’re raising an interesting question that if I’m putting instructional materials that I’ve created on a third-party site, as opposed to my own hosted site, I’m actually transferring rights to that site. That is a really scary question. I would say the answer is yes, just because vacation pictures aren’t the only thing that social media companies care about claiming as their property as well. But as a general question to conflict of university ownership, you could make an argument that the university can only license to themselves what’s in their system, and therefore there’s no way for them to use what’s in a third-party system because they have no access to it. So if you provide students links to external sites, which may include third-party sites, I think you can get around that licensing issue because it faces like an efficacy problem where it’s just simply not feasible for the institution to get access to it. It wouldn’t take very long for their lawyers to draft new policy that would say you can’t use third-party sites for our classes because even though I don’t think FERPA would have any bearing on this, you could just simply call it a FERPA risk issue. You could call it anything that a lawyer would want to call it. So they could marshal those back into the institutional LMS. They could also say that you must provide them copies of what you create for your courses out of a courtesy of making sure the curriculum is consistent. So they could get around that, and then you’re back to where you perhaps have started. So that could be an issue down the road. But I think initially, certainly on your own created site, you would maintain all ownership over those materials. And if you give students access to those materials, again, you’re not transferring copyright to them. In a way, you’re almost giving them a limited license to have access to your materials, not really sharing ownership of it with them.

John: On a related note, suppose that you’re writing a book and you share drafts of it with your students in an LMS. Are you giving up some of your property rights to that material if the course was created as a work for hire?

Kent: That is a super scary question, and I would say that generous heart of saying “let me share some materials with you” might inadvertently cause you to sort of step on your own rake head there and find out that you got slapped with the realization that you’re losing ownership rights over that material. If it’s protected by copyright… I have put drafts of books that I’ve written, chapters or parts of chapters, into an LMS to share with students as well as an alternative to maybe the book I choose for the course and say here’s an alternative way of looking at the material because that’s already owned by the publishing company. I’m not granting them any rights because I don’t have rights to own it as well. But unpublished work that you maintain the copyright in because you own it as soon as you write it, putting it in an LMS in a way at least grants a license to that, unless you figure out a way to scrub it from your LMS at the end of the semester and remove it as an archived piece of information.

Rebecca: AI has been a very popular [LAUGHTER] topic on this podcast and just in general, and folks are certainly utilizing generative AI platforms to edit materials, enhance their work, use it as a dialog for feedback, and those kinds of things. So, does the user own any rights to images, text, and other content that they’re using generative AI to create?

Kent: Boy, here we go. Who can take another AI discussion, right? It’s like when the zeitgeist has become full tilt, and we just can’t take it anymore. But this is relevant, and it really is a needful question for faculty to think about. So, if I can just do a background on AI and give you a lawsuit involving an Indonesian monkey. So a British photographer named David Slater became friendly with some monkeys in Indonesia as a wildlife photographer. He gave them names. One was named Naruto. He gave them his 35 millimeter camera, and then they started taking pictures of themselves. He then published a book of monkey selfies, which is charming in a variety of different ways. And in came people for the ethical treatment of animals, PETA, who on behalf of this monkey named Naruto sued the author of the selfie book in American court, alleging he violated the copyright of the monkey who took the pictures. I wish I could make that up. I lack that creativity. The upshot of that lawsuit, as it relates to AI, and it does, is this: unlike Solomon, everybody lost. PETA was thrown out of court in effect for saying you have no standing to bring a lawsuit on behalfof a monkey in Indonesia. The photographer lost because he didn’t take the photographs, so he didn’t have the copyright in them. Now you might say, “Well, didn’t the monkeys take them for him?” No, they’re not his employees, so it’s not a work-for-hire monkey selfie book he was working on. Poor Naruto lost as well, ultimately because of this. Copyright law is clear. To be an author, that doesn’t mean just for written word, that means for anything that’s copyrightable. to be an author, you must be a human being. Now, by way of contrast, corporations are persons under the Constitution, but the word “person” is not the word “human being.” So, to be a copyright author, you must be a human being. That moves us a few years later to AI. Nothing that comes out of large language model AI is copyrightable. Now, if we get to this, it’s to me copyright infringement, but it’s not copyrightable because a computer is not a human being. So no matter what it is, if it comes out of generative LLM AI, it is not copyrightable. We know of this for two reasons. One, there’s an interesting case where a super smart guy created his own AI system and had it create some sort of impressionist style painting that he gave a title to, I think it’s called “A Recent Entrance to Paradise,” and he sought a copyright registration for it. It was denied. He went through the court system as far as you can go, and he lost every time because AI is not an author and it can’t be. He tried a second bite at the apple by saying that the computer system he generated was a work for hire, and he owned the work itself. And the court’s like, you know what? Again, work for hire start with a copyright author, that has to be a human being. Now that means we know that from case law, and the Copyright Office in 2023 issued a very long pamphlet statement, which clearly states this: there is no copyright in AI-generated work at all. Here comes the however, and there is one, and they did say this, and there’s a case on it. They said there could be, on a case-by-case basis, a copyright granted to what the copyright office called AI-assisted work, not AI- generated work. I would just say, by a quick analogy as an example, if I ask AI to write an essay for me, or let’s just make it worse, let’s make it a paper, let’s go there. I ask AI to write me a paper I want to get published, and then I take what they give me and then I edit it myself. Not only is that unethical and should be unpublishable, it’s not copyrightable because that came out of AI.

But suppose I write an essay that I’d like to get published on today’s educational issues going on in America, and the requirement is that it must be 2500 words, and I just can’t get down from 2700. If I put that into the AI system and ask it to get me to 2500, I would argue that I still own the copyright in that because I used AI to assist me in a relatively minor editing method, as opposed to, in effect, co-generate with me. So case-by-case basis for AI-assisted work, but AI-generated work there is no copyright whatsoever.

Rebecca: I’m so glad you shared some of these stories. These are some of the cases I like to share with my art and design students. [LAUGHTER]

Kent: Well, can I do an art design for you? [LAUGHTER] Because the case I’m thinking of is… and I’m terrible with names. The name has slipped me, but a person used AI to write a graphic novel. I’m old enough to call them comic books, but I guess that shows I’m too old to not know they’re graphic novels. But a really interesting looking graphic novel in which AI was used to generate all the art, and when that author of the graphic novel saw the copyright in it, it was denied because it was AI-generated art, and then the author sought a rehearing, you might say, of that, and they got a Solomon-like outcome in which the copyright office said, “Okay, here’s the thing: you own the words that are in the bubbles in your graphic novel, and you own the placement of them with the AI-generated art, but you do not own the AI-generated art.” So that means if I wanted to use any visual from that graphic novel, I am free and clear to do so because nobody owns it. But I can’t use the cells because the cells are copyrightable, and certainly the language inside those thought bubbles are copyrightable as well.

John: What if you write something and then you run it through AI and it makes some suggestions for change and you adopt those suggestions? Does it depend on the extent of AI assistance in terms of that case-by-case analysis, or would it be determined on where the thoughts originated, whether they were human originated thoughts that were then cleaned up and edited by AI or rewritten or restructured. It seems like there’s no clear dividing line between the work of the human and the work of the AI.

Kent: That’s why, and this sounds like a cop out, and I apologize, but that’s why case-by-case basis. While I’m saying I don’t think the copyright office was wrong in to say that, it creates a line that you’ll never see the end of, of finding out have I done enough for that to work. Like in corporate law, shareholders can be personally liable for the debts of the corporation if the corporation is undercapitalized, but that’s not based upon a mathematical basis of examining debt to equity. It’s done on a case-by-case basis. So, in your scenario that you gave as a hypothetical, I would say to me that sounds like that would belong to the human being who started with the work. But how much of that goes from doing fine tuning to actually creating really what is a new thing with the use of AI? When it becomes a new thing, as I read that copyright advisory document, it’s really not protected by copyright. And the scary thing to think about is this, and I know that some AI companies will tell you they’re still not doing this. You believe them at your own peril. But once you put your own work into an AI system, in my view, you’re contributing [LAUGHTER] voluntarily to the copyright infringement that is the reason for the existence of AI in the first place, and so the web scraping that goes on to create an AI system, which is the subject of a variety of large copyright infringement lawsuits that are still ongoing, you’re at least volunteering your work in an AI system that may allow your work to then be used for training and other purposes that maybe you don’t like as well. So there may be systems that claim that you can eliminate that, or prohibit that, or stop that from happening. Again, the next question asked is how would you take that to court? Because we live in a world that’s controlled by arbitration agreements, and if you have a membership or a subscription to a paid AI system, they’re not going to want you to sue them. So just like with your cell phone agreement and your credit card agreements, you’re going to have to arbitrate that, which would be a nightmare anyways. And so in a way, it’s almost like I would say when they say don’t lend money to family, treat it like you’re giving it away. I would presume to say the same thing, if you’re going to put your work in an AI system, presume you’re giving it away for purposes you can’t control. But if the purpose why you need to use it in an AI system overwhelms that, then just go ahead and do that.

Rebecca: So I want to like pull a thread that you’ve hinted at in a few different places already, which is some of the work that we may have created that are our own intellectual property has been scraped and is part of these systems. And this is certainly something that comes up a lot in my land of art and design, where particular images of artworks or whatever may be scraped and now are part of these large language models that are now used to create these images that are not copyrightable, [LAUGHTER] right? And so faculty may be using AI to generate images and things to illustrate their ideas or other things, but it’s pulling on a database of things that don’t necessarily belong to them. Can you talk just a little bit about that angle?

Kent: Yeah, the question. might be, if I’m hearing this correctly, is it possible that a faculty member is infringing on the copyright of somebody else when they’re using AI materials that they’re pulling out, even if they’re strategically asking for what they want to ask, because what has gone into AI is, to me, an infringement of someone else’s copyright. And I think that’s a big question because even though there’s no copyrighting what comes out of it, what went into it, in my opinion, has violated the copyright of the author. And so I would say that faculty aren’t infringing on the copyright of an AI system by using AI for whatever they’re using it for. The lawsuits that are ongoing are still ongoing, and there have been some settlements. I’m an author of a textbook that’s owned by Cengage so I know I’m in a settlement with the AI system Anthropic, and that settlement is in, I think, the billions of dollars. But what it gives is to each copyright holder a couple thousand dollars. And I think it’s kind of unfortunate, but that is the way that it’s going, is that the capital that those AI companies have is so huge that they can settle a class action lawsuit. And some authors have said, “I’m not going to be a part of this. I’m going to sue individually.” To them, I say, “Godspeed.” But those settlements allow them to just simply wipe that as an expense and then continuing what they’re doing, but the ethical issue is: Am I doing something I really should be doing? Am I using AI to generate ideas for myself or quick answers I can’t find, or am I using it to actually create work that I want to claim as my own if I massage it with my own assistance when in fact it came originally from another author or even an academic author who sort of lost their rights to it when it was web scraped into an AI system. It’s a deep question. It’s not a copyright infringement to the AI system, but it could be sort of an ethical infringement to the original person who didn’t choose to lose their work into an AI system.

John: Now let’s take that just a little bit further. Suppose you used it as a faculty member to generate some imagery that goes into a book that you sold to a publisher, and then the imagery that was created by the AI platform involves a copyright violation of the original copyright holder.

Kent: Wow.

John: Who would be legally responsible for that violation? Would it be the AI platform that generated that, or would it be you as the author who is using essentially slightly modified copyrighted material?

Kent: Yeah, that is not only novel and really brilliant. It is, to my knowledge and my research, unanswered as of now, so I’m only projecting what I think is the case, and I do see a world where perhaps an original owner, let’s say, of art that ended up in an AI system that ended up again in someone else’s work, could sue that end user, saying you infringed on my copyright, and you could find a situation where a faculty member might have to ask themselves, “Do I have an insurance policy that would cover the legal costs of defending myself in an infringement lawsuit?” Because let’s say that suppose that happened to me, perish the thought, is that a situation where Purdue is duty bound to defend me? But if, like you said a book, what we’ve said is these books tend to be not owned by institutions because their copyright is exchanged for some fee structure. But if you read your publishing contracts carefully, you owe the publisher a hold harmless clause to make them whole if they violate copyright because of what you put into a book they published, and so that’s a very scary financial question. And again, that’s something to think long and hard about before using AI-generated art. And again, it doesn’t matter if you have that background knowledge like Rebecca does. It matters if you’re simply using work that really is sourced to somebody else, because the middleman, in this case AI, is not liable for it, at least not yet, and maybe not ever. But you, the end user, could be.

Rebecca: We haven’t talked a lot about the role that students play. Students are often creating work; they’re submitting it to you, maybe as a faculty member to review. Now you have a copy of it in your possession, perhaps. Do students own their own classwork?

Kent: There is a lot to say about students. It’s not just the cheating part, although that is worthy of discussion if you want to talk about cheating and AI detectors and things. But students own their work for the same reason that faculty might own their work as well, and I think students should be treated respectfully with respect to their work. And so, if we say that the student work passes the modicum of or spark of creativity test we talked about, that does belong to them. There can be some exceptions. Sometimes institutions have policies for graduate students if they’re creating work that’s part of a project that is sponsored by the institution. Sometimes there might be policies that say students don’t own their work if it has been hired out by an outside party. For instance, suppose a marketing class has an instructor who’s connected to a business that wants a project done that the students can free help the business, right? Well, who owns that work is a huge question. Policies might carve exceptions out there, but as a general basis, students own their work. So I have, in instances, particularly in online classes, said to students, “You know, some of your work here is so good. Is it okay if I use it anonymously as a template or a model of what ideal student work would look like?” I shouldn’t take their work without their permission. If I take their permission, I seek an email that I can archive and say they granted me permission. There is a case, there are more than them, but there is a case out of a university in Detroit where an art student was asked by her instructor to have students create artwork that would be used by the Mercedes-Benz exhibit at the Detroit Auto Show, and she knew what she was giving up, but she didn’t realize that Mercedes, with the knowledge of her instructor, used more of her work that she created than she had acknowledged to give over to Mercedes. And she sued Mercedes-Benz for copyright infringement, and she won. And it didn’t matter that she didn’t have a market value in her work because there’s another part of copyright law that allows you to get statutory penalties for simply infringing on someone’s copyright. And then she sued her professor for what is commonly known as a breach of a fiduciary duty to her as a student. I don’t know where that lawsuit ended up, but the court allowed that lawsuit to go forward. So I would say that faculty should be mindful that what students create in their coursework, including artwork, musical compositions, and just essays and projects, really do belong to the students, absent some external independent contractor work for higher situation like the Mercedes Benz was for three of the four artwork pieces that ended up at the exhibit. So I do think faculty should be mindful of that. And there’s a separate issue, which is a faculty member could say, “Yeah, but there’s no fair market value in this student’s work. So what does it matter?” Well, that question alone shows why it does matter, because faculty should be respectful of their students as individuals and as owners of their own work. So whether there’s a market value for it to me is not the question. It is, “Does it belong to them?” And if it does, and it does, then I should be respectful and just ask to use it. And if they say yes, get it in writing somewhere, and you’re good to go generally,

John: Following that up a little bit, if we upload student work into a plagiarism or AI detection software package, or perhaps just to an AI platform to get some feedback on student work to assist with grading, would that be a violation of students’ rights? I was thinking in particular of how if there was some artwork in that student’s work that was uploaded that was used for training it, it would be in some ways a similar sort of violation.

Kent: I totally agree with that. Often AI and students, from an HR standpoint at work, is discussed with respect to FERPA. Well, I mean, you’ve got to be an idiot to use AI systems to create education records that would violate FERPA. But in terms of what you’re saying here is, am I really transferring without their even knowledge, perhaps, their ownership into an AI system by using it for evaluation purposes? That is a serious thing to consider. Now, would anybody know that I did that? They could argue the only way that my work ended up in an AI system is I turned it into you, and you stuck it in there. So it wouldn’t be terribly hard to figure that out. But that’s an ethical question and a legal question as well. One issue to think about with respect to AI detection, if I can just mention that, is when you’re using it, you really then transferring copyright because the original plagiarism detector that we all knew… I never used it… in fact, I didn’t pronounce it correctly unless someone corrected me on it is Turnitin, but students from a high school class sued Turnitin for copyright infringement, and I say “you go, good for you,” and they said you took my papers and violated my copyright by sticking in to Turnitin, and the students lost because the court said this is a transformative use of someone else’s work for plagiarism detection, and there is no infringement on their ability to publish their work by being archived for plagiarism detection purposes. That’s not, however, what’s going on now. Because even Turnitin’s AI system has been turned off by some universities. So I would say this: here’s some free legal advice, which means it’s worth nothing. Don’t use AI detectors. Simply don’t do it. It’s not worth it for a variety of reasons, and the most important reason is to me it’s error prone, and the last thing a faculty member should do from a legal protection standpoint is accuse a student of cheating based upon an AI detection system because they have been known to be error prone, they’ve been proven error prone, and even if they’re improving through retraining, they are error prone. If I haven’t said that yet, and if you’re saying I’m giving up my sort of autonomy as a judgment maker to an AI system, which is based upon predicting the next logical word, that’s why they fail. That’s why they make mistakes because they’re based upon perplexity, “What is the next most logical word coming up that we think you want?” So if I say I’m dropping my children off at blank, logically it’s going to be perhaps school or the playground or the pool. That’s the lower perplexity. Well, you’re giving up your autonomy, your judgment decision when you’re saying to AI, check if this has been cheating. Now, if you can reverse engineer that by saying I put your work in an AI system and it showed that it came out of already published work. That’s a traditional plagiarism detector. But AI systems also flag often non-native English speakers as cheating because of the way they write sentences with low perplexity and what’s called burstiness, which is an interesting word. Burstiness is the variation we use in the sentences we write: short, long, those kinds of things. So, non-native English speakers and sometimes autistic students are flagged because of the low burstiness in their sentence structures as being cheating. So, to me, just don’t use AI detectors because AI detectors also sell humanizers to students as well. AI humanizers are like police officers selling radar detectors on weekends. So if you want to avoid being caught by us by this product, when you’re dealing in a world where AI humanizers are already being used by students to bypass your using AI detection, then the question is, what are you doing for your job? Is that your job to be the hall monitor of all hall monitors? And the cheating is rampant. Just recently, I think a Brown University award-winning economics professor realized that allowing his students to do a take-home exam basically the AI cheating was rampant, and he was just horrified by that. So we know that it’s going on, and I’m not saying that we shouldn’t care about it.

What I’m saying is, out of all the time you have in your life, of which a lot of it is work, how much of it can you give to hunting down the possible AI cheaters in your classes? I don’t have that kind of time, and if I’m going to be wrong, I’m really wasting my time as well. So I think there’s a legal issue with using AI detectors. MIT, who you can, if you want to, you can source MIT as the first creator of an AI system, says don’t use them. So if MIT’s teaching and scholarship learning center says don’t use them, why do I need to seek a second opinion? I’m not going to use them, and so I’m not saying you’re wrong to use them. I’m saying you use them at your own peril because of the issues involving the last thing I want to be accused of is committing an act against a student that is known as arbitrary or capricious. Sometimes it’s called arbitrary and capricious. One of the few areas where a faculty member faces a legal risk is when they act arbitrarily towards a student, and I believe using an AI detector as your reason for accusing someone as cheating is getting you into that arbitrary world that I don’t think is worth risking.

Rebecca: On the flip side, students often submit tests, assignments, and other materials created by the instructor into generative AI platforms [LAUGHTER] without the consent of the instructor. How should faculty address this issue?

Kent: One, and I know you don’t need me to say this, so I apologize for preaching to the choir here. But have a syllabus policy first of all, because this may seem too smart by half, but a student has a good point by saying I didn’t do anything wrong because you didn’t say I couldn’t. And so if we presume they’ll know they shouldn’t do something, then we’re wrong as being in charge of our classroom. So, have a syllabus and specifically delineate what you don’t want to be done. You, again, can’t control what they’re doing on their own time, and unless you get back to that Paleolithic world where we had everything done in class, either on paper or in front of people, it’s hard to catch those kinds of things. People sometimes have students sign honesty pledges attached to syllabi. Those are all well-meaning. They have no legal value. So, other than policies that allow you to go to the dean of students or have some sort of institutional disciplinary process for that if you catch it, outside of a clearly delineated policy and examples and encouragement. And again, this may sound hokey, but like I tell my students, you’re in a business school. A business school is a professional school. What does it mean to be professional? Like, what does it actually mean? It’s not about making money. What does it mean to be a professional in a professional school? How does ethics fall into that? And again, that may seem like I’m wasting their time if they don’t care, but at least they know that I’ve thought about it ahead of time, and that I tell them it’s not worth getting caught, and you’re not going to feel better down the road because you cheated here, and so I just try to be that good angel on the one side of their shoulder, reminding them that it’s just not worth it. Again, that may be a waste of my time to tell them that, but I can’t hunt down all the possible cheating unless I bring everything back into the classroom, which some faculty are again trying to do, and we’re now living in a like a record player world again, where we’re saying, you know, something, let’s go back to the record player and let’s bring the classes not just back into the classroom, but bring the assessments down to more of that low technology world, which cuts down the chance of cheating.

John: We always end with the question. What’s next?

Kent: Well, I’m thinking about or trying to work on creating a podcast for myself, not to compete with you at all, but on legal issues for faculty, and I’m trying to work out how that might come together, and hopefully that will come to fruition sometime in the near future. And I’m working on some conference presentations to interested faculty members and book groups where we get together and discuss issues from this book that I’ve written on legal issues for faculty, and then the fall semester is around the corner, so that’s on its way as well.

Rebecca: Yeah, summer never fails to slip by.

Kent: No, I finished my summer semester about a week ago, and now I’m thinking about the fall semester already.

John: And it’s always good to see more podcasts related to higher ed, including legal issues in higher ed. We’ve lost a few good ones in the last year, so it’s nice to see new ones coming online.

Kent: Well, if I can overcome the technology difficulties I had prior to this recording, that may be something to figure out. Can I actually make this happen? So, thank you for the encouragement on that.

John: If you need any help, just let us know. [LAUGHTER]

Kent: Well, that’s recorded now, [LAUGHTER] so you can’t walk that back. Rebecca didn’t say it; you said it. So I have one person to go to, and I can pinpoint when you specifically said it. So thank you for that invite. [LAUGHTER] Il I appreciate it more than you might care to know.

John: Well, thank you. It’s great talking to you. And these legal issues are ones that many faculty are thinking about. So this was really helpful.

Rebecca: And I always enjoy these conversations. So thanks for joining us.

Kent: Thank you very much for letting me be here. I really have a wonderful time talking with the two of you. So thank you again for allowing me to be on your podcast.

[MUSIC]

John: If you’ve enjoyed this podcast, please subscribe and leave a review on Apple Podcasts or your favorite podcast service. To continue the conversation, join us on our Tea for Teaching Facebook page.

Rebecca: You can find show notes, transcripts and other materials on teaforteaching.com. Music by Michael Gary Brewer.

[MUSIC]