Dr. Andres Guadamuz on IP in the AI Age: Exploring Copyright, Authorship, and Future Regulations

Show notes:

1:20 background and work in IP law and technology

2:20 blog article, “What can internet history teach us?”

3:25 IP issues emerging in the age of AI

5:10 inevitability of AI

6:40 global regulation of AI

8:25 Emily Gould - which body would handle global regulation of AI

11:00 Council of Europe’s adoption of first AI international treaty 

11:50 Gould - UK proposal to expand text and data mining exception to cover commercial uses 

14:55 transparency issues 

18:40 Gould - response - need for legislation

20:20 authorship question 

21:40 THJ Systems v. Sheridan (THJ Systems Ltd. v. Sheridan [2023] EWCA Civ 1354, [2024] E.C.D.R. 4, CA, 20 November 2023) is of great interest because it confirms the test for originality in copyright law in the UK after Brexit.

22:55 Li v. Liu, Case Number: (2023) Jing 0491 Min Chu No. 11279, Beijing Internet Court, 27 November 2023

25:00 NFTs

25:30 Thaler v. Perlmutter and USCO, USCA Case #23-5233 

 29:30 continued utility of copyright 

32:40 AI copyright suits in the US

36:30 cultural impact of AI models’ accelerated training capabilities 

38:20 view of whether there is a future for careers in art

42:50 tools like spawning.ai for artists’ protections 

43:00 opt ins versus opt outs

43:50 technological protections like Glaze and Nightshade 

45:15 difficulty of implementing opt ins 

47:45 injustices in the AI age and definition of justice

50:33 mark that Andres hopes to make with his work

53:40 Stefania Salles-Bruins - IP protection for AI software and outputs

Please share your comments and/or questions at stephanie@warfareofartandlaw.com

Music by Toulme.

To hear more episodes, please visit Warfare of Art and Law podcast's website.

To leave questions or comments about this or other episodes of the podcast and/or for information about joining the 2ND Saturday discussion on art, culture and justice, please message me at stephanie@warfareofartandlaw.com. 

Thanks so much for listening!

This podcast and its content may not be used for training or developing AI systems without permission.

© Stephanie Drawdy [2026]

Transcript

Dr. Andres Guadamuz 0:01

I mentioned this in a recent blog post. May you live in. Interesting times is a curse and, my God, we are living in interesting times.

Stephanie Drawdy 0:13

Welcome to Warfare of Art and Law, the podcast that focuses on how justice does or doesn't play out when art and law overlap. Hi everyone, it's Stephanie, and that was Dr Andres Guadamuz, a reader in intellectual property law at University of Sussex and editor-in-chief of the Journal of World Intellectual Property. What follows is a recording of a recent Second Saturday Art and Justice online gathering, during which Dr Guadamuz and I are joined by Institute of Art and Law's Emily Gould and artist and attorney Stefania Salas-Bruins to discuss IP in the age of AI. Dr Andres Guadamuz, welcome to Warfare of Art and Law and Second Saturday. Thank you so much for being here.

Dr. Andres Guadamuz 1:08

Thanks for the invitation. I'm delighted to be here.

Stephanie Drawdy 1:11

If you would, could you give a bit of background and what brought you to focus on IP law and technology?

Dr. Andres Guadamuz 1:21

Well, I've always been a geek, since very early, a gamer, sort of a geeky guy, always liking technology and computers. I used to build my own computers. So whenever I went into law it just became natural to sort of concentrate in sort of the techie aspects of law, because it was just I didn't have formal training but I always found it quite interesting, so sort of I gravitated always towards that. So initially I was very interested in copyright from the very start and I always found it quite interesting and also technology in general. So usually software and software development and open source software I'm not a developer myself, but I was always interested in this and that sort of led. One thing led to the other and that's where I decided to specialize and do all of my area of specialty.

Stephanie Drawdy 2:26

Yeah, and the blog that you've maintained, for it seems like it's been for a few decades now, I think. Is that right?

Dr. Andres Guadamuz 2:36

Yeah, it's going to be 20 years. Someone else told me this and I hadn't really thought about it. Yeah, it's definitely going to be 20 years in October, I think. So yeah, I sort of started blogging for a while.

Stephanie Drawdy 2:54

Yeah, it's a trove of information, and I just reread your most recent post about the timeline of like from the golden age into the AI age. It was a very comprehensive overview of the progression to where we are now and I wonder if you might give your thoughts on some of it. You've touched on in your blog the concerns or the interesting issues that you've seen that have been currently arising in this AI age.

Dr. Andres Guadamuz 3:28

Yes, I've been interested in sort of AI also for a long time. I think it was sort of early One of the people who was interested in the legal aspects quite early. I think that it's always fascinated me how potentially, the law is always going to have to respond to technologies in general, that sort of has been the driving force of my scholarship and I've ended up thinking of technologies or looking at technologies that may shape society quite strongly, Some that haven't had as much an effect as some would say. Blockchain, I think, potentially, is one.

Dr. Andres Guadamuz 4:16

Some people think that we're still early, but I've always been a skeptic and in regards to AI, I've always been fascinated as a geek. I'm also a science fiction geek. I've always been fascinated by stories about artificial intelligence in general, robotics and Asimov's loss of robotics and all of these things that have a very interesting regulatory interest. So that's where I've sort of gone with artificial intelligence. My emphasis or my analysis has been sort of to look at what's going on specifically with the area of IP and AI. So where things are headed, I cannot tell you. I have very strong ideas but I don't know if I'm right at all.

Stephanie Drawdy 5:07

Would you like to share some of those very strong ideas.

Dr. Andres Guadamuz 5:13

I think it's probably something that people sometimes don't like to hear, and it's that it's a little bit inevitable.

Dr. Andres Guadamuz 5:21

I just don't see it's one of those technologies the closest thing I can think of is the actual the internet itself. So I think that there is a lot of hype, for sure, but it's going to be incorporated in our lives. It's already a big part, and we don't tend to think of things that we already have as AI, like Google Maps and Apple Maps and all these applications that tell you where to go, almost like magic you end up in a place and how long you're going to take there. That is, in a way, an implementation of AI, and photography and Photoshop and things like this are already quite heavily using AI. So I think that, starting from the point that it's inevitable and I know not everyone thinks this and I am very happy to discuss this but starting to that point, I am fascinated in looking at how the future is going to be shaped and where things are going to head, and what regulatory approaches can we take to sort of ameliorate the obvious problems that everyone can see, problems that everyone can see.

Stephanie Drawdy 6:42

Just touching on the regulatory aspect for a moment, since you raised it. What are your?

Dr. Andres Guadamuz 6:53

thoughts about the need for, and the potential of, a global approach. This is a little bit of a tricky one. I'm a big believer in treaties and harmonization, but I know how difficult it is. So there is my sort of internationalist, idealist person, optimist that I am. I would like to see more. More of an international treaty or things like the EUAI Act, which was just published, are steps in the right direction, I think. But also I am concerned that some of these harmonization efforts are very difficult to get off the ground. Just to give an example, it's been 20 years, or over 20 years, to get a WIPO the World Intellectual Property Organization to negotiate a treaty on traditional knowledge, and that's a very non-controversial subject. So I find it difficult to believe that we would probably get any sort of international AI agreement. I think potentially the EU effect, or the Brussels effect, as people call it, that a strong bloc like the EU passes a regulation like the GDPR a few years ago and now the AI Act, that has a cascading effect on regulation across the world.

Stephanie Drawdy 8:26

We've been joined by Institute of Art and Law's Assistant Director, Emily Gould, and so I'll hand the floor over to her for a couple of questions.

Emily Gould 8:35

Thanks, Stephanie. It's just a quick follow-up question to yours. Hi, Andrew, it's really good to meet you. I follow your writings on this and many other topics and they are always so insightful and interesting, and so it's great to get to meet you. I follow your writings on this and many other topics and they are always so insightful and interesting, and so it's great to get to meet you. Very briefly, Just to follow up on that question about whether there should be or could be some kind of international approach to the regulation of AI. What, if anybody, do you think has the kind of authority or the interest in pushing that forward? Would it be WIPO or could it be something like UNESCO or the UN? Ip aspects is the only candidate.

Dr. Andres Guadamuz 9:15

Unesco is a little bit of a strange institution. I'm saying that my dad worked for UNESCO, so I'm a UN brat.

Stephanie Drawdy 9:35

I grew up in different countries.

Dr. Andres Guadamuz 9:39

So it's an institution that is very dear to my heart, but its power to shape international policy is diminished. I think there are political issues surrounding that. During the Cold War it was seen to be left-leaning, at least in US policy. It was sort of left outside and therefore it's great for sort of exploration and talking and having reports. I think I've read a couple of really good reports from UNESCO, but, yeah, as a norm-setting body I don't see it.

Dr. Andres Guadamuz 10:27

The UN itself there has been some movement there, there have been a couple of reports by the UN Secretariat on this subject, but again, I don't see this as something that would potentially lead into some form of agreement. It's very difficult for the UN now to sort of generate agreements in that way, and that's just politics. Again, it's sort of very difficult. So our best hope for some form of UN-led regulation could be, precisely, could be WIPO. There is a Council of Europe which they gave us, the Cybercrime Convention, which has been important. Parts of that have been adopted around the world and I think that they are. They've actually produced the text very, very quickly. So that's one to keep an eye on. I don't know how influential it's going to be, if it's going to be as influential as the Cybercrime Convention, which, again, it's Council of Europe, but it has been approved by or implemented by lots of countries.

Emily Gould 11:48

Interesting, interesting. I think getting agreement generally, whether on an international level or national level, is just so very difficult. I was just thinking of one example in the UK, which you probably know about, which we probably know about, but, um, so there was an attempt by the government, well, a proposal by the government, to expand the text and data mining exception to cover commercial uses.

Emily Gould 12:14

And then, um, then there was a great outcry by the rights holders, uh, and you know various collecting societies, and so there was a complete about turn. And then they said, ok, well, we'll look at codes of practice, voluntary codes of practice. And then they set up this working group and it met quite a few times, I think over 10 months or so, and then they just announced whenever it was spring this year I think it was that the two sides were just simply unable to come to an agreement and that kind of put an end to that. And it's just a very general sort of agreement to continue talking about the topic and to look at, you know, transparency. But it all seems to have sort of died a death, those attempts to create some kind of a code, even on a national level.

Dr. Andres Guadamuz 13:03

Yeah, I was looking that closely. Um, I was not invited. So I don't know, uh, I was, uh I was never invited to participate in that, but it's probably because I was very vocal that it was never going to work. Maybe I thought it was doomed from the start. I'm I'm very pessimistic when it comes to certain things like that. I'm very optimistic normally, so I think I always thought that that was doomed to failure. I thought that the government had to produce some form of top-down exception. I wasn't a fan of the government's first proposal. I thought that was way too broad. I'm rightly criticized by everyone. I'm more a fan of the European approach, so I think that the government could have just gone. Okay, we'll just do the same thing as Europe is doing. Also, maybe thinking I'm a remainer or rejoiner now.

Dr. Andres Guadamuz 14:16

I guess just makes my life much easier and just thinking in the future towards harmonization with Europe. It would just make more sense. The idea of the UK government at the time with this code of conduct was precisely to sort of one-up Europe in some ways, sort of try to say oh yes, they have an opt-out, so we're going to have a system that doesn't have an opt-out. That means all of the AI companies are going to come here. So it was misguided from the start. It was absolutely and rightly. It didn't go anywhere.

Stephanie Drawdy 14:57

On the transparency point that was referenced a bit ago, I wonder if I could get your thoughts on the transparency issues that might be arising and how you think they could be addressed.

Dr. Andres Guadamuz 15:14

I think that it's going to be quite interesting field the transparency. Now, the EUAI Act is going to have transparency requirements. Providers or people who are creating models are having this transparency requirement, so they pretty much have to disclose what copyright works went into training. We have no idea how that is going to work, but that's a conversation for another time. However, that is in itself quite interesting because I think that the AI industry, looking at it from their perspective, they're very concerned about transparency, from what I've been hearing, and it's because everyone that has been transparent is getting sued. So it's pretty much you show what's in your model and the lawsuits follow. So it's in their best Sort of. The people who were very early transparent were the first ones that were the recipients of lawsuits and that sort of. It's a good thing in many ways, but it's also a bad thing because actually it makes transparency makes the industry don't want to be transparent and it's in their best interest actually to want to be transparent and it's in their best interest actually not to be transparent, when I think actually transparency would work in everyone's interest, potentially for what I think is going to be the future, which is going to be a combination of exceptions, limitations and particularly licensing agreements, and I think that we are going to find an equilibrium, because that's how things always end in this area.

Dr. Andres Guadamuz 17:10

You mentioned before that I'm a big believer in internet history. We have to look at internet history and every new development has been followed by lots of lawsuits and then, at some point, there is equilibrium. At some point, either regulators step in or sort of the case law becomes very clear. So if people get sued, they know that the case is going to get thrown out of court. So I'm thinking user-generated content, technologies like YouTube, which have been allowed to continue.

Dr. Andres Guadamuz 17:45

Web browsers, which pretty much have been allowed to continue. Web browsers which pretty much have been allowed to continue, even though, when we think about it, web scraping is a large copyright infringement. File sharing, which was pretty much obliterated out of existence through lawsuits, and intermediary liability, in which intermediaries platforms were allowed to continue and be ensured that we're not going to get sued out of existence by things like copyright infringement that were committed by their users. So all of these eras have always been followed by strife, but also, most importantly, equilibrium. Some form of compromises are reached, et cetera, et cetera. So I think that's where, eventually, we're going to be heading.

Emily Gould 18:39

Yeah, I think it's interesting how, with each new, new sort of major technological development, you do have that period, as you say, of sort of strife and lawsuits and nobody knows quite which way things are going to go, and the courts have to do an awful lot of work and there's often there's often this sense that oh, it's a big wild west out there, you know, anybody can do what they want, which is of course never in fact the case, because there is always some law to fall back on, some case law, some analogies that can be made.

Emily Gould 19:07

And I was listening to one of the the leading IP judges, lord Justice Arnold, the other day talking about.

Emily Gould 19:15

He was actually talking specifically about copyright and computer generated works and how that's sort of a murky area and it's been brought much to the fore, um at the moment in the context of ai, and he was saying that at the moment, you know, the courts are sort of doing their best and we're already seeing lawsuits, um, but there will come a point where, um, he seemed to say, and this was a sort of big conference about ai and you know lots of different talks and the general kind of tenor of the arguments tended to be, I think that, um, you know, there will be this period where we get, uh, we get court judgments, but there will ultimately need to be some legislation, um, covering some of this new stuff, albeit that, you know, um, many of the arguments are already dealt with in some way, shape or form, by existing law.

Emily Gould 20:04

There will need to be some kind of legislation, and that sort of makes sense, particularly, I think, in the UK, in the context of the EU Act. It feels sort of strange and that it couldn't work long term for the UK to be completely out of step with what's happening just across the channel.

Dr. Andres Guadamuz 20:21

Yeah, I absolutely agree. The authorship question is in itself extremely interesting. I've sort of been a proponent, a very early proponent, that some works that are generated by an AI should have copyright, sort of I stuck my head out very early on that, but it's mostly for practical reasons. More than anything else, I think that at some point we're going to have to come to the recognition that lots and lots and lots of works, and increasingly more and more works, are going to have some form of AI into the process, be it editing, be it some form of manipulation of images, effects, all sorts of things, and we're going to have to come to terms with that and, I think, with that understanding we're going to potentially have to have rules in which you know if there is some intellectual creation that went into the creation of the work, that can be just pressing buttons and setting things.

Dr. Andres Guadamuz 21:36

There is a very interesting case recently THJ I think, and Sheridan I think, from last year which was very, very basic stuff that was done to a computer program like a graphic user interface and it was considered to. This has copyright. It is extremely basic. The action or the intellectual creation or the intellectual labor that went into the creation of this was very low, but that doesn't mean that there wasn't any, and I think, as an avid user of AI myself, I can see lots of differences. There are some things that I can see. I put a lot of effort into this. If I hadn't used Photoshop, why shouldn't I get copyright? Yes, the machine ended up doing some of the things, but actually there is usually quite a lot of thought, and not only the prompts, but also the selection of outputs, the selection of inputs. You get some outputs and you start refining things and I and we get we give people copyright for less yeah, yeah, there's been that interesting case in china as well, hasn't there?

Emily Gould 22:58

internet court, where they're talking about exactly all of those things, even, um, even sort of choosing the particular ai tool in the first place, you know, an element of creation, and yeah yeah, and and people who are really good at this.

Dr. Andres Guadamuz 23:13

Um, I, uh, I know quite a lot of people in the ai art community. I know probably you're not supposed to call it ai art. I think ai generation or whatever you want to call it Prompting generating I think it's another term that people use, where people produce some horrible stuff, some really poor things that have no effort whatsoever, and some people who have really gotten very good at the technology and they put in the hours and you can tell in the results that the results are sometimes very interesting, very thought provoking. But that's the same with every type of creative industry that we've ever had. The same happens with photography.

Dr. Andres Guadamuz 24:00

My selfie is Probably my selfies are not very good, let's put it that way, but you know they are in some ways a result of my intellect, even though the machine does nowadays more of the work. But some people have brought photography to an art form and it's rightly. You know it's a show of skill and we recognize that this is sometimes very good, and I see the same. There are people who are doing fantastic, very interesting work and there are people who are using it to produce a very low-level anime girlfriends that are absolutely trite and horrid to look at. But that's poetry, you could argue. It's the same. I'm not a poet, but we recognize that quality is not a requirement for protection.

Emily Gould 25:02

It was very much the same with NFTs, wasn't it? There was some amazing work coming out and some fairly low-level work, yeah.

Dr. Andres Guadamuz 25:10

Oh yeah, don't get me started. I was fascinated by that from a legal perspective.

Emily Gould 25:14

Yeah, it's not.

Dr. Andres Guadamuz 25:16

Yeah, I found it extremely interesting, even though I wasn't a fan of most things, but, yeah, there were some people that were producing quite interesting stuff. Again, it's like with everything in life producing quite interesting stuff again.

Stephanie Drawdy 25:32

It's like with everything in life. This prompts me to uh raise the case uh brought by the inventor, stephen taller against the us copyright office. I'm curious what you uh, if you have a thoughts about that case and the us uh human authorship requirement.

Dr. Andres Guadamuz 25:47

Yes, I would like to start by saying that I love Stephen Thaler. I'm not just saying that he's a godsend for IP lawyers who specialize in AI. I think that all of the lawsuits that he's been brought they have brought my classes to life. We have lots of fun. I don't know. I've joked before that I have no idea if he's an AI sent from the future to try to sorry science fiction geek, what can I say? It's sort of interesting how much money he's spending. Just let's get that out of the way. But the lawsuits themselves, I think that he's been.

Dr. Andres Guadamuz 26:36

The problem is that they have been a little bit ideological in the sense that he's trying to get an AI to be the inventor itself or the author itself. So the first applications listed the machine or the AI or the model as the author in the case of the copyright office, and I think that's problematic because it was never going to pass. Never going to pass. The argument has shifted in the appeals where it's oh no, the author is Stephen Thaler, but the machine was only a vehicle. I still think that it's an interesting one. I can see why it lost. I think it was just argued on the wrong part. I would probably say that it's been an interesting exercise.

Dr. Andres Guadamuz 27:43

Now the US Copyright Office has had a very a more, much more reluctance on recognizing AI authorship. I think that this is not going to last. If I am honest, they're already granting copyright to some works that have been generated by an AI where there has been a substantial amount of human input into it. So the last time I heard, they had issued hundreds of copyrights and this has sort of gone a little bit under the radar. So it's already happened that they have granted copyright to AI works. My guess is that sort of a little bit on the quiet, more works are going to continue to have copyright that at some point were generated in one stage or the other using AI, and there's also part of my practical answer on this. The practicalities of this question is that it is going to be very difficult for us to try to determine where AI has been used in every case.

Dr. Andres Guadamuz 28:56

I've mentioned it a couple of other times that it's like there is no copyright police knocking down on doors and looking for AI use, and I think it's just practical to sort of recognize that if there has been a substantial amount of human intervention, whatever that may be, that the word should have copyright.

Stephanie Drawdy 29:20

Also just the overarching question of whether or not copyright has perhaps been demoted in its utility. What are your thoughts? I know I've heard you speak about this, but would you share your thoughts and your current view?

AI Art, Copyright, and Artist Protection

Dr. Andres Guadamuz 29:37

Yeah, this is a tricky one because I think that potentially, ai, if we start being very picky on what gets to get copyright protection, one of the reasons why there is no registration almost nowhere else in the world is precisely to try to make the barrier of entry very low to copyright protection entry and very low to copyright protection, and I think that potentially, that could have a detrimental effect to copyright in general because people are not going to bother. Something that I've noticed quite a lot in AI-generated or AI art communities is that they don't care about copyright. They really don't. So a lot of the debates that we were having before are almost academic, because what happens is that they produce a work and what happens is with the community, there is very little monetization or sometimes there is a little bit of NFT use, which is copyright is absolutely irrelevant when it comes to NFT. What you want is a smart contract that tells you that you're the owner of this digital file, or at least you own the provenance of this. Now that has led to most people who are doing this for a living or using this every day. They couldn't care less about copyright.

Dr. Andres Guadamuz 31:17

This is sort of in line with a community or some communities that I've seen emerge online, like the creator communities, the influencers communities online. Like the creator communities, the influencers communities, content creation in general which are communities that are more worried about copyright because they're going to get copyright strikes from YouTube or from Instagram or from TikTok, and they are less concerned about the actual copyright of whatever they produce, because they are live, they're being monetized by entirely different ways and they don't sell anything that is copyrightable. It's a live stream and they're the providers and the economy is entirely different. So there is a very interesting creator economy that is emerging where copyright has nothing to do. It is like copper. The only inference of copyright is sort of a bother oh, I'm going to get demonetized. So for these communities, ai is just another thing that they can use to produce faster content. Faster content means more subscriptions. More subscriptions mean more money.

Stephanie Drawdy 32:31

Turning then, with this copyright question in mind, to the 20-plus lawsuits that have been filed just in the US. You had mentioned earlier that transparency was the hitch that got many defendants in court. There are several complaints, I've read, where one of the allegations is that the defendants, the AI developers, have not been transparent. So it's kind of interesting that you would say that where AI developers can go from here, depending on how suits go like would they just be going to other countries that are more friendly and leaving the States? What do you think?

Dr. Andres Guadamuz 33:14

Yeah, a fantastic question. I don't know what's going to happen. But in regards to transparency, the first lawsuits I was referring mostly the first lawsuits were to the companies that have been transparent early on. So a big example of this is Lama, the model by Microsoft. They had published a couple of papers where they had said which datasets they had used. So people that were contained on those datasets were some of the first claimants. The same with the early OpenAI models, particularly, I think, gpt-2. There were papers that had said we are using this to train REI, and so early on they were very transparent and those transparency reports or papers were used by the claimants.

Dr. Andres Guadamuz 34:27

What has happened is that they are not disclosing anything that they've used anymore. That practice of transparency didn't survive the first few lawsuits. So, yeah, what's happening now is that there is more obscurity of what is being used and I think that's potentially the later, particularly the later models are. I mean, the later lawsuits are saying okay, we are guessing that our words were used to train the AI because of this, because there is a similarity, it's producing outputs. It is, there's a similarity, it's producing outputs, there's a similarity there that we have to assume that our works are being used, and this is where I think potentially the transparency thing is going.

Dr. Andres Guadamuz 35:25

As to the second part of the question of whether this is going to potentially mean that developers are going to go elsewhere, yeah, the answer is potentially yes, particularly because we've only had a handful of lawsuits outside of the US. So, just on average, it's easier for you if you're training your model somewhere else, particularly the EU. I think it has the potential, even with the EU AI Act, to be considered sort of training-friendly jurisdiction. There are others, potentially China, even though there are some lawsuits and Japan, so we don't know, but I'm expecting that also, some training is going to go to countries where there is almost no existent copyright enforcement. So that could be also the case, but I'm not sure if there is going to be a migration, but there is always a possibility.

Stephanie Drawdy 36:31

One of the concerns that I've heard from artists about having had their work included in data sets is this idea that AI models have accelerated training capabilities compared to humans and they don't mind sharing their work with humans to collaborate, but not, at least without their consent, to share them unwittingly with AI models. Do you have thoughts on the cultural impact that kind of shift might have where AI models are producing at accelerated rates this work, that, whether it's identical or not, it has some influence? That's coming from artists in the data sets, yeah.

Dr. Andres Guadamuz 37:16

I think that artists are completely justified in being worried about this. If I was an artist, I would probably be worried, worried. It's a tricky one, because I think that some of the concerns are guided in the sense that, or are right in the sense that there are serious concerns about livelihoods, about replacement. I think that some artists are going to have quite a lot of trouble earning a living in the future. There is no other way of saying that or nice way. My niece is an artist. She's sorry to mention her, but we had a conversation recently and she loves art and she's always drawing, and she was asking well, uncle, do you think there is a career in art? And I said, no, I'm sorry, do it as a hobby? And I had to say, I think I had to be honest, I had to say, like, sorry, I don't think that there is a career in art. I'm sorry to say this, however. So I can see that and I also see quite a lot of concerns regarding consent and I think it's very important the issues of consent that artists are seen to be giving their consent of whether or not their works are going to be used. Having said that, I think that there has been a little bit of an effort to push certain narratives that are sort of doomsaying a little bit. So this is definitely going to be affected, but I think that there are also potential opportunities in the future.

Dr. Andres Guadamuz 39:27

And I keep thinking back of music in the file sharing era, which is an era where musicians knew, oh my god, this technology is going to kill us, and musicians were rightly worried about file sharing and piracy. And eventually this was true. It bore out that there were quite a lot of very legitimate concerns on music. Now some musicians panicked and started suing everyone left and right and some musicians sort of adopted very interesting music sorry business models. They tried to be more proactive, be more active with their merchandising, recognizing that the age of the $20 CD was dead. So that model was going to be dying, but new opportunities arose. So music nowadays is pretty much free. If you think about it. What musicians are making their money is in different ways with all of the surrounding things, and I can see potential opportunities in the future for art.

Dr. Andres Guadamuz 40:48

I think that there is going to be a hunger in the future for human-made art. There's absolutely. We're going to be flooded. We're already being flooded with so much bad AI art. Just browsing Facebook is like filled with absolute horrible things. So I think that definitely, there is going to be a little bit of a lot of opportunities.

Dr. Andres Guadamuz 41:17

I've joked sometimes that there is going to be like a business model for proof of humanity or something along those lines where proving that you're a human artist is going to be very much a selling point, and I think that artists already should be trying to place themselves instead of all the sky is falling, let's all run to the hills and to be more proactive in this sense of thinking. Okay, there's a big challenge where there are potentially also big opportunities for people who take either adopt AI which I know it's also something that some artists are absolutely against but also for those who are not interested in adopting it to become sort of a brand of human artist Only I don't know. I'm thinking of sort of organic food or things like this. All sorts of certifications are going to arise in which you have, you can prove that your work is 100% human made, or things like that. So there is concern. By the way, I try always to say this very clearly People are justified in being concerned.

Stephanie Drawdy 42:38

And perhaps there is a potential for your niece to have a career in the arts.

Dr. Andres Guadamuz 42:42

Yes, exactly, human-made art by Alejandra Guadalos.

Stephanie Drawdy 42:51

What might be your recommendations for artists looking to protect themselves their work? Spawning AI is one that comes to mind.

Dr. Andres Guadamuz 43:00

Yeah, spawning has been great, and not only because they interviewed me a few months ago. I think that what they're doing is great. Go, I think that what they're doing is great. Those types of opportunities I think that potentially are very good. Trying to opt out I think eventually we're going.

Dr. Andres Guadamuz 43:24

Some people advocate for opt-ins. I think opt-ins are never going to work out. It's just so difficult and I know potentially, ethically opt-ins would be the best way, but practically it's different. So I think that the way to go is going to be opt-outs, and things like spawning are a great opportunity for that. Yeah, I think opt-outs is the main one.

Dr. Andres Guadamuz 43:51

There have been a few technologies that are being advertised Glaze, nightshade and some sort of technological solutions. I have to admit I'm not a big fan of them because I think that they're very easy to circumvent. I'm not a technical guy, I'm just a geek. I've seen people claiming that they're easy to circumvent, so I don't know if that is true, but from my perspective it looks quite a lot of work, quite a lot of very intensive work. That potentially makes it more difficult to produce art for something that potentially is not going to give you protection. So it is a tricky one, unless the technology changes, I don't see how that could work, and also historically, technological protection measures do not work. Pirates or people who want to circumvent things find it easy. We found that out from the file sharing years as well. Dvds came with protection. It's still very, very easy to circumvent.

Stephanie Drawdy 45:10

Interesting that your thought is the opt-in would be too difficult. I kind of want to hover on that for a moment if we could. Is it too difficult? Because it would slow down technology training too much.

Dr. Andres Guadamuz 45:28

No, it's just difficult to implement. Opt-ins are always extremely, extremely difficult unless you have an ironclad type of regulation that requires everyone to opt-in and only models can operate with with opt-in. I just don't see how it would work and Because without that, no technology is going to, no company is going to try an opt-in system, maybe some sort of. Again, there is going to be a niche market for sort of organic, sort of ethically produced AI type of thing. There is definitely always going to be a market for that, but again, the same as with food, it's a niche market. So for mass production I just don't see it happening from a practical perspective.

Dr. Andres Guadamuz 46:26

I'm very pragmatic maybe that comes across, I try to be, and that sometimes gets me into trouble because I say, well, what you're saying is horrible, andres, you should be emphasizing opt-ins. And I said, yes, ethically I can, absolutely. I am all for transparency, I'm all for opt-ins, but sometimes I just I have to be honest and I don't think that some of these things are going to work. I'm more also looking at how we're going to end up. I'm a big believer in sort of the big picture, the what's after the AI wars, and I just don't see opt-ins working. I honestly unless you get all of the companies to agree on an ethical framework, which they're never going to do either. We live in a capitalist society, which is another question. I'm a big socialist.

Stephanie Drawdy 47:40

This seems like a good segue for asking your thoughts about a question I usually raise with guests about justice, and do you see injustices or potential injustices in this current AI age, and what is your view of definition of justice? Has it evolved for you?

Dr. Andres Guadamuz 48:05

Oh, yes, that's way above my pay grade.

Dr. Andres Guadamuz 48:11

I've always liked to think that I'm for justice and whatever definition you want to have of that, and definitely there are quite a lot of injustices when you think about inequalities particularly.

Dr. Andres Guadamuz 48:31

I think that the way we are going in some ways, is going to just be the entrenching of business models that are extremely unjust, unfair, whatever you want to call it, and one of the things that I've always been a proponent of is openness and open source and open access, open systems, open licensing, and I think that's sort of an antidote to all of this in which we can have I know this word is overused, but democratization of content, which is sort of the ideals of Creative Commons, the ideals of open source software in which more people are able to produce things, and I think actually there is a potential there for AI to be a force potentially in that direction.

Dr. Andres Guadamuz 49:27

But also that is going to be accompanied by lots of injustices. And what I'm also afraid, particularly a very, very big concern of mine, is that we are sleepwalking into a system in which, whether it be through regulation or licensing, we're going to end up with the same tech giants or different giants, or giants that we're just to end up with the same tech giants or different giants, or giants that were just being born right now and they're going to be the new capitalist powerhouses of the future, where only those can comply with regulation and that is always a very big fear of mine can comply with regulation, and that is always a very big fear of mine, because, potentially, if you regulate only for Facebooks and Googles and Microsoft of the world, everyone else gets left behind, you cannot get started, and that is also, in a way, a big injustice. That makes sense.

Stephanie Drawdy 50:31

Yeah, All of this work that you're doing and thought that you've put into an effort for this analysis of the intersection between IP law and tech, what's the mark that you hope to be creating with this analysis that you're doing and archive, really that you're creating?

Dr. Andres Guadamuz 50:53

I don't know. I like to have fun. That's my goal in life is to have fun doing things that I love. I love researching, I love writing. I write for free. Every single word that I've ever written is free online, so maybe that sort of gives you an idea of where my philosophy is coming from. I've never made a penny from anything I've written and, yeah, because I enjoy people reading my work and that's really it. It's mostly that I just enjoy and enjoy and having fun doing it. I try to put fun into everything I do. Doing it. I try to put fun into everything I do. Linus Torvald of Linux fame once said why do it if it's not fun? Or if it's not fun, why do it? So I've always used that as sort of my motto. I try to always have fun whenever writing or doing things. That's really what I like, and if people read my thoughts and find them interesting, that's even better. That's the only reward that I seek.

Stephanie Drawdy 52:18

And I will say I haven't noticed, if you've put it on your blog, whether or not you create the images that go with each post. I assume that you do, and they're quite entertaining and they reek of fun.

Dr. Andres Guadamuz 52:32

Yeah. Yeah, I've stopped sharing a little bit because, again, there is a little bit of a cringe element when it comes to AI images. Nowadays it's just so much bad images now that AI generation of images has become a little bit tarnished, but I still would like to think that I'm quite good at it, at least some of the images, yeah so, at least the images I try to always and it's a fun thing that it has helped me quite a lot my blog post before I would just find an image on Google and I would inf reflect what I want to do with a blog post.

Stephanie Drawdy 53:12

Yeah, so, in that respect, it has been a great help. Yes, yes, the techno llamas.

Dr. Andres Guadamuz 53:37

Yes, lots of llamas, of course, lots of llamas.

Stephanie Drawdy 53:41

Yeah, stefania, as we wrap up the hour, did you have any questions for Andres?

Stefana Salles Bruins 53:47

Yeah, I do. I don't know if I'm going to open a can of worms again, but thanks for all the information. I've mostly been on listening mode learning, and I was wondering if you could touching back on a topic you referred to earlier about lawsuits trying to get IP protection for the AI, for the software or maybe it's the developers behind that software that would be protected? Like, what would it take to go that route and do you think we'll go there? You know, I think you touched on it a little bit.

Dr. Andres Guadamuz 54:41

Yeah, I definitely think that that's where we're heading. Just from again, from a practical perspective, it's going to be more difficult to try to discern whether or not something has been generated with AI that it hasn't, so why bother? It's sort of I see it like what's happening with marking, with grading right now in universities. We are assuming I'm assuming that my students are using AI all the time. I can see it lots of times. You can either try to pursue it, but it's very difficult to prove or just grade the work as it's presented to you, which is much easier. And if it's a bad essay, it's a bad essay regardless of who wrote it. Otherwise it would be impossible. I would spend all of my time trying to pursue.

Dr. Andres Guadamuz 55:29

So I think that's where we're heading, just from a practical perspective, that it's just going to be easier to live our lives with the assumption that most things are going to have some AI at some point. So that's where I see things are heading. It's just less philosophical and more practical, more pragmatic. Now, eventually, what's going to change is that everyone is potentially going to be using it all the time. You know there are people who will say for sure that they're not, but I think that potentially that's where we're heading. There is going to be some form of AI, either copy editing or checking images or improving some images. So yeah, I don't know if that answers the question.

Dr. Andres Guadamuz 56:21

I think, it's just just like again, I guess exhaustion, tiredness.

Stefana Salles Bruins 56:28

To follow on that. Is it conceivable from your point of view that the software being used and the people behind that software being used so openly like you're describing, will kind of turn and want protection for that output?

Dr. Andres Guadamuz 56:47

Yeah, eventually, I think it's just what's going to happen. At the moment, people don't care, as I mentioned, but what's going to happen is when more and more commercial enterprises and commercial products are going to be produced using AI. I suspect already that lots and lots of things that we're seeing on screen or scripts I would make a joke about the Acolyte and Star Wars being produced entirely by AI, but I won't take a cheap shot Star Wars being produced entirely by AI, but I won't take a cheap shot but I would suspect that already we start to see things on screen art already produced, videos already produced, music already produced that is highly produced with AI. I would not be surprised by that, and I think it's just. People are just keeping it quiet. So I think that potentially, what's going to change things is when Disney starts producing AI-generated cartoons or media. In that respect, they're going to lobby to have this protected by copyright.

Stefana Salles Bruins 58:12

Once the big boys start using it for commercial purposes.

Dr. Andres Guadamuz 58:16

Yeah, and for commercial purposes is where this is going to be tested entirely. At the moment, it's mostly hobbyists or people like me who are using it for non-commercial uses, Like the images. At the moment, it doesn't have a commercial value. When it does, that's where the meter is going to be moved. I'm not sure if I'm using the right expression.

Stefana Salles Bruins 58:43

Yeah, thank you, that was interesting well.

Exploring the Future of Technology

Stephanie Drawdy 58:48

Thank you so much. Thanks, stefania, for the questions, and thank you, andres, for being here, for answering all of our questions and sharing your insight. It's been such a pleasure. Is there anything else that you wanted to share that we have not asked you? Oh no, my god, I think we've covered pretty much everything. Is there anything?

Dr. Andres Guadamuz 59:04

else that you wanted to share that we have not asked you? Oh no, my God, I think we've covered pretty much everything. I think it's just an admonition that the future is interesting.

Stephanie Drawdy 59:18

There will be links in the show notes to learn more about Dr Guadamuz and his blog, maintained under his online persona, technolama. If you are intrigued by this podcast, it would be much appreciated if you could leave a rating or review and tag Warfare of Art and Law podcast Until next time. This is Stephanie Draughty bringing you Warfare of Art and Law. Thank you so much for listening. Roddy bringing you Warfare of Art and Law. Thank you so much for listening and remember injustice anywhere is a threat to justice everywhere.