Insights

When AI Invents

JONES DAY TALKS®: When AI Invents: Two Applications Test U.S. Patent Law

Two recent patent applications seek protection for inventions that were created autonomously by artificial intelligence without a human inventor. The applicants want the AI to be deemed the inventor and the AI’s owners to receive the patent rights.

Emily Tait and Carl Kukkonen discuss how the United States Patent and Trademark Office is exploring the relationship between AI and patent law.

Podcast: Play in new window | Download

SUBSCRIBE TO JONES DAY TALKS

Subscribe on Apple Podcasts
Subscribe on Android
Subscribe on Google Play
Subscribe on Stitcher

LISTEN TO PREVIOUS PODCASTS

Read the full transcript below

Dave Dalton:

Artificial intelligence is growing more powerful and gaining application in many areas. AI can now create new innovation on its own without a human inventor, a capability that will only expand as the technology progresses. But whether patent protection should be available for innovation invented by AI remains an open issue. We have a Jones Day panel here to discuss the underlying legal issues and explain what might be next. I'm Dave Dalton, You're listening to Jones Day Talks.

Dave Dalton:

Emily Tait has served as a trusted advisor in complex intellectual property disputes and transactions, for matters involving diverse technologies, including vehicle software, healthcare data, fintech, medical devices, telecommunications, pharmaceuticals, biotech, and consumer products.

Dave Dalton:

Carl Kukkonen has more than 20 years of experience in strategic intellectual property, counseling and litigation. He advises clients on patent infringement and validity, the preparation and prosecution of patent applications, pre-litigation case assessment, patent litigation, IP due diligence and brand protection matters. Emily, Karl, thanks for being here today.

Carl Kukkonen:

Thanks Dave.

Dave Dalton:

All right, let's set up some context for this discussion. We'll go to Emily first. We looked back at how intellectual property rights were structured here in the United States. They were meant to promote innovation by protecting the interests of artists and authors and banners as they brought new works into the commercial world. But historically, and implicit in all of that was that authors and inventors were assumed to be human beings. Is that correct?

Emily Tait:

Yeah, that's absolutely correct, Dave. And when you think about just the word intellectual property, it's the property of one's intellect and intellect has historically been understood to be something that's uniquely human. And so certainly the US constitution and our statutes governing IP contemplate that innovation will be created by human beings, either inventors of patented technologies, authors have created works under the copyright laws, et cetera. And so we've had a variety of developments in recent years that have sort of challenged that paradigm. And you'll remember a few months ago when you and I talked about the monkey selfie case, that's one example.

Dave Dalton:

By the way people, it took us a whole minute and a half to get to the monkey selfie story. But go ahead, Emily. Sorry.

Emily Tait:

Right. So the monkey selfie case, of course related to a case that went all the way up to the Ninth Circuit Court of Appeals, dealing with whether an animal in that case, a monkey has a legal right to be an owner of a copyright under US law. And the court there said no. And the copyright office has clarified indeed that there was a human authorship requirement. So a monkey can't be an author of a copyrighted work and nor can a machine where there is no human intervention. And so that really begs the question of how much human intervention is going to be required under the copyright laws.

Emily Tait:

And then of course we've had this development recently relating to patented technologies. So we saw this with the monkey selfie case that the case may have implications beyond just monkeys and indeed for artificial intelligence. And that happened rather quickly with respect to patented technologies.

Dave Dalton:

Sure. Well, let's get into that a little bit. We'll go to Carl for a second and Emily started to allude to this, but the possibilities brought about by artificial intelligence, which were unimaginable probably 10, 15, 20 years ago, certainly, but that could upset the whole apple cart in terms of what's been the norm in terms of patent protection. What is going on here with AI, creating innovations? What are we literally talking about?

Carl Kukkonen:

So artificial intelligence is being applied across numerous industries, and it seems only natural that it's now being applied to the innovation industry. With AI, at some point a human operator needs to define an objective for the AI models, whether it's to find a new drug discovery target, or whether it's to increase the efficiency of a process, or whether it's to give a credit score for a consumer, AI has been very valuable in making these processes more efficient as people get more and more data. It seems only natural that it's now being applied to innovation and that the human operator can define a model and have a data set of some sort, which might be previous patents, for example, and use that data set to then create a new patentable subject matter.

Dave Dalton:

Yeah, well you mentioned processes. Even as a lay person, I'm not a lawyer, I'm certainly not a scientist, but I can understand how AI can adapt to a process and make things more efficient and better and more accurate, perhaps, et cetera, but innovating creating, this is a big leap forward, this is just a side of science fiction. Isn't it? I mean, did you see coming?

Carl Kukkonen:

I did not see this coming but I think AI is becoming increasingly autonomous as well as ubiquitous. And so it's adaptive, it can grow and maybe like the Stephen King movies from the eighties or the Terminator series, AI does has this ability to kind of again, be autonomous and also tackle challenging problems.

Dave Dalton:

This is one of those areas where our technology, if we're not careful might outrun our ethics. I guess I'm getting off the track here. This is a stunning area with a lot of developments to come I'm sure.

Dave Dalton:

And we talk about the two patent applications, and that's why we're here today. This is reflected of a commentary that you can find at jonesday.com. The title here is; "When innovation invents." I guess the issue here is there were two patent applications filed by the University of Surrey in England. Talk about what actually happened there, Emily.

Emily Tait:

Yeah. So a team of individuals from the University of Surrey filed patent applications in the United States Patent Trademark Office, as well as a number of other patent offices worldwide. And this team identified an AI system called DABUS as the named inventor. And DABUS is an acronym standing for the Device for the Autonomous Bootstrapping of Unified Sentience. But that AI system is actually listed as an inventor. And so what basically happened is the human researchers at the university developed DABUS to generate certain ideas and then to determine which of those ideas were sort of the best of the bunch. And then DABUS would essentially "invent these inventions" without human intervention. And that's what the University of Surrey team says happened here. So an invention for an improved beverage container and an improved neural flame device used for search and rescue missions. And they're saying that that inventive step was performed without human intervention whatsoever and by the AI system itself as the inventor.

Dave Dalton:

So they've applied for patents, but the patent office hasn't acted yet?

Emily Tait:

That's correct. I don't know the exact status in the patent office at this moment, but certainly the patents have not been granted. And so there's a lot of discussion right now about, what's going to happen with these because it really challenges the existing paradigm where you have human beings as inventors.

Dave Dalton:

Okay. So let's talk about the rationale here. So Carl, what is the university's position that patent ownership should extend to the owner of the AI? Turn the AI loose on this project? And we set the agenda, we put in the parameters, so whatever AI creates belongs to us. Is that their take?

Carl Kukkonen:

Well, I think that the university researchers are doing this to create a discussion about how these types of innovations, which are AI driven are going to be treated by the various patent offices. It raises many different issues with regard to even just patent office procedure. For example, we have an obligation to identify all the relevant prior art that's being used or considered by the inventors in connection with the patent application. And so in this case, do we need to have some sort of an explainable AI model, which talks about everything that the AI model use to innovate in a certain area. So how do we get that information? How do we provide that to the patent office? That also raises issues that maybe Emily can talk about with regard to litigation and how do you typically investigate the activities that were taken in connection with the patent application?

Dave Dalton:

Why don't we talk about that. Emily what's your take there?

Emily Tait:

Right. Well, so in this entire discussion, I agree with Carl, that these applications were filed to sort of spark dialogue. But it's really, if AI can be an inventor, it disrupts our system of patenting and therefore would massively disrupt the way patent litigation is handled in terms of trying to conduct a deposition, right. You're trying to investigate with the inventor, how the conception happened and how do you do that when the inventor is AI. So there's just...

Emily Tait:

The possibilities really are limitless. And there are so many more questions at this point than there are answers, but these applications really are sparking that dialogue. And indeed the USPTO has been front and center trying to get this dialogue going, dating back, even at least a year. I know earlier this year, there was a focus group at the USPTO discussing artificial intelligence innovation. There's been a variety of comments made by the director of the USPTO on this point. And most recently there was a solicitation of comments, explicitly related to artificial intelligence inventions, seeking public comment on a variety of issues that are really core to this discussion.

Dave Dalton:

And as we speak today, there's still soliciting comments. People can still chime in, I hope?

Emily Tait:

Yeah. And so this notice came out in late August a few weeks after actually the patent applications were filed by the University of Surrey. This notice from the USPTO, and it sort of describes the situation and some of the uncertainty surrounding artificial intelligence inventions, and then ask for comment by October 11th on a variety of different questions relating to AI innovation. And so October 11th is the deadline. The comments will be made available to the public for inspection. So we'll shed light on what different industry participants are thinking about these core issues.

Dave Dalton:

Okay. Let's go back to Carl. How might this ultimately play out? Is this the kind of thing that could take a lot of say years? I mean, this is complicated. The people that introduced patent laws and regulations and intellectual property rights, years ago, they couldn't envision something like this? This is a whole new paradigm, I think. Is this something that takes a long time to work itself out and could it finally maybe ultimately require action by Congress? What do you think?

Carl Kukkonen:

It's likely that legislative action is going to be required to resolve these types of issues. The Supreme Court has heard many cases over the years, going back, even as early as 1980 in the Diamond v. Chakrabarty case, in which the Supreme court found that a human living organism was patent eligible subject matter because the Federal Congress, their intent was that anything made under the sun by man is patent eligible. And I guess the question here is whether invention, which is driven by an AI model is something that's made by man. And maybe a more gender neutral term is going to be appropriate for these days.

Dave Dalton:

Let me ask you this. And I hope this doesn't sound too cynical. If this gets to Congress, lot of hardworking, good people there, do they have the knowledge background, the understanding of patent law, and certainly an advanced technology like AI to do the right thing? I know they've got staffers, they'll get input, hopefully they'll do some research, but is this something the Congress is equipped to handle, I wonder? I don't know who else would do it by the way, but this is a whole new thing. Any takes there?

Carl Kukkonen:

I think Congress has done a good job historically of getting input from various stakeholders, including for example, IPO, the intellectual property organization, pharma and others, as well as other lobbying organizations, which advocate for intellectual property rights owners. And I think once these organizations find some sort of a consensus internally, then they're going to be advocating their positions with Congress. So I think it is likely, and maybe with the input that the director gets from the public as Emily mentioned earlier, we may start getting some guidelines on what's going to be patentable. What's not patentable. How do you define inventors and maybe how you address some of these other issues that arise.

Dave Dalton:

So when the time's right, there'll be ready, it sounds like. That's great. Let's wrap up with this. What are you telling clients right now? How do they plan for AI based innovations in how they protect their interests? Let's go to Emily first.

Emily Tait:

Yeah. I think clients right now just have so many questions. It's a tough scenario because there really are more questions than there are clear answers under the current law. I think it's important for any type of client that is utilizing AI, developing it internally or acquiring it, obviously to be very conscientious in how that's done. So if it's developed internally having clear indications and documentation to reflect how that development took place, if it's being acquired from a third party, conducting diligence to ensure that they understand how that AI was developed by the entity that they're acquiring, so that ownership issues are clear and so that they have answers to questions as far as how these systems were developed. At this point, I think the human nexus is significant because our current system, certainly for copyright and also for patent contemplates human participation and indeed authorship and the copyright context, having this human connection to the AI development is important.

Emily Tait:

And I think a final point is just encouraging clients to pay close attention to these developments because artificial intelligence is so hot right now, not only with the US Patent and Trademark Office and the copyright decision, but also it's a priority of the Trump administration. It's a focal point of the government, as far as staying competitive in a global economy. And so there's going to be a lot of discussion going forward this year and beyond on AI. So paying close attention to that, I think is a very significant point for clients.

Dave Dalton:

Stay informed. Carl, what would you add?

Carl Kukkonen:

So my observation is that clients are increasingly trying to get patents on their AI inventions. Patents are a good vehicle, very strong vehicle. Some of the inventions are not always patent eligible. And so they need to be careful about what they're trying to protect via that route, via other routes. Trade secrets may be the most beneficial approach for many clients, because it may be not necessarily the models that they're using, but it may be the data sets that they're using or the way that they're using the outputs of their models in their processes, which end up being more important. And so in different words, it would be to consider other types of protection in addition to patent protection for their AI related innovations.

Dave Dalton:

Good enough. Now, Emily I'm imagining the way these things usually go, probably the PTO is seeking public input and comment on its approach to AI. If that's the case, how does that work?

Emily Tait:

Yeah. So in late August, the USPTO issued a request for public comments on patent related issues relating to AI inventions and sought public comment on a variety of issues and asked that comments be received by October 11th. Then in late September, they bumped out that deadline to November 8th. So there's a bit more time for members of the public to weigh in on this. And these are really fundamental questions. So things like, what are the elements of an AI invention? What are different ways that a human being can contribute to conception of an AI invention and be eligible to be named as an inventor? So really going to the core of inventorship do current patent laws and regulations regarding inventorship need to be revised to take into account the scenarios of where an AI invention may wholly invent something or meaningfully contribute to an invention.

Emily Tait:

A variety of other topics related to obviousness, for example, does the standard of what would be obvious to a person of ordinary skill and the art need to be revised to reflect AI. So does AI impact the level of a person of ordinary skill? And if so, how? So a variety of questions, but really just seeking public comment. And as I said the deadline has been bumped out from October 11th to November 8th to give interested members of the public, an opportunity to weigh in, and the comments can be provided by email. So AIPartnership@uspto.gov and those comments will then be available for public inspection as well. So it's a really interesting process. And the USPTO is very much engaged in the topic and looking to see what industry participants are going to say on these key issues.

Dave Dalton:

Sure, okay. And we'll make sure all that information you just gave is on the jonesday.com website and is tagged on this podcast. I'm always curious, who responds to these things? You said people in industry, people from other government agencies, maybe academic types? I'm wondering who weighs in. Because there're so many different stakeholders potentially. Based on your experience, who are they likely to hear from?

Emily Tait:

Yeah. I think there's going to be a variety of stakeholders who have really specific and pointed views of how these things should be handled. But I think there will also be stakeholders that hold back, not wanting to sort of show their hand or perhaps they don't have a definitive view yet on how these things should be handled. So I would imagine that there is going to be a range of participants just as you said, government entities, perhaps various agencies, companies, corporations, academics, and perhaps even lay people. There may be sophisticated corporations down to an individual inventor who's thinking about these issues, engineers who work in artificial intelligence, who may have strong views of these things. So I think it'll be interesting to see who ultimately weighs in and who decides to sort of watch from the sidelines and see what others say on this. Obviously it's going to be an ongoing going dialogue.

Dave Dalton:

Yeah. Well, it's funny. I was about to say what you just said. I thought you'll get a lot of people hanging around and say, okay, what's everybody else saying, what's obvious, what's less obvious where can I chime in. So I think it's a fascinating process. So Emily, thanks

Emily Tait:

Yeah, I think that's right. It'll be interesting to see what happens from here because there's just so many open issues not only in the patent side, but AI and trade secrets and AI and copyright, and also even just AI's impact on trademarks. So there's just such a huge discussion right now about AI's impact on intellectual property across the board that, you know, this is an important step taken by the USPTO, but obviously it's just not a final step by any means. It's getting a conversation going forward in a meaningful way. So we will be watching it closely.

Dave Dalton:

We will leave it right there. Emily, Carl, thanks for being here.

Emily Tait:

Thank you so much.

Dave Dalton:

Fascinating area of the law. I'm sure we'll be talking again soon. So as this develops, stay in touch and we'll talk soon. Thanks so much.

Emily Tait:

Thank you.

Dave Dalton:

For complete bios of Emily and Carl, visit jonesday.com. And while you're there, be sure to check out our Jones Day Insights page, where you'll find podcasts, publications, videos, newsletters, and other compelling content, and look for the famous monkey selfie podcast featuring Emily and Jones Day talks regular, Meredith Wilkes. That was a classic.

Dave Dalton:

Subscribe to Jones Day Talks, apple podcasts, and wherever else podcasts are found. As always, thank you for listening. I'm Dave Dalton. We'll talk to you next time.

Speaker 1:

Thank you for listening to Jones Day Talks, comments heard on Jones Day Talks should not be construed as legal advice regarding any specific facts or circumstances. The opinions expressed on Jones Day Talks are those of lawyers appearing on the program and do not necessarily reflect those of the firm. For more information, please visit Jonesday.com.

Jones Day publications should not be construed as legal advice on any specific facts or circumstances. The contents are intended for general information purposes only and may not be quoted or referred to in any other publication or proceeding without the prior written consent of the Firm, to be given or withheld at our discretion. To request reprint permission for any of our publications, please use our “Contact Us” form, which can be found on our website at www.jonesday.com. The mailing of this publication is not intended to create, and receipt of it does not constitute, an attorney-client relationship. The views set forth herein are the personal views of the authors and do not necessarily reflect those of the Firm.

 
We use cookies to deliver our online services. Details of the cookies and other tracking technologies we use and instructions on how to disable them are set out in our Cookies Policy. By using this website you consent to our use of cookies.