Artificial Intelligence: The Growing Role of AI on Patents
The emergence of artificial intelligence-related technology as a means of innovation has led to uncertainties for companies across industries, primarily because U.S. patent law has historically held that intellectual property rights be assigned only to humans. Jones Day's Matt Johnson, Carl Kukkonen, and Emily Tait describe the current state of the patent law landscape as it applies to AI-originated innovation, and explain the importance of implementing an effective strategy for protecting key IP assets.
Read the full transcript below:
Carl Kukkonen:
There are certain areas of AI that don't fit squarely within any of the categories that we currently have. For example, a training data set. So a training data set is going to be used to inform a machine learning model, but the data set itself may not be patentable. It's maybe not subject to copyright. It could be maintained as a trade secret. And one of the things that people have been talking about in the industry is that there may be a need for a new regime of intellectual property to protect aspects of AI.
Matt Johnson:
Normally, an invention is just something that someone working on R&D or out of the blue comes up with in their head. Now companies are using AI as a tool to help innovate, and even in some cases using it to potentially innovate on its own. And that's a real shift in the paradigm from what we've done in this nation for 200-plus years where this patent statutes have said that an inventor is a person. Well, now what's actually doing the inventing is a computer, it's artificial intelligence. And how do we adapt our system to handle that?
Emily Tait:
AI has been a topic that the USPTO has talked about, certainly in a very significant way over the past year. And in the summer of 2019, and then the fall, the USPTO issued two notices via the Federal Register, where they were seeking input from the public and from stakeholders in this area to say, "Look, we have all these open questions on AI-related technology and what it means for patents." So for example, if the artificial intelligence system itself helps conceive of an invention, what does that mean? Because under the current system of laws, a human inventor is contemplated. And so the first notice addressed this issue on patents and AI's impact on patents, but then there was a follow-up notice that said, "We have all these other aspects of IP. So how does AI impact copyright, trade secret, databases and even trademark law?"
Emily Tait:
And so those notice periods have now ended, the USPTO has accumulated many, many comments from many stakeholders, and they are in the process of evaluating those comments. And the current sort of announcement from the USPTO is that they will probably get some type of guidance in the spring that perhaps summarizes those comments and offers some preliminary thoughts on the impact of AI on IP.
Carl Kukkonen:
I think legislation is probably where we need to end up with regard to defining what is patentable or what can be protected under copyright or other regimes. And also identifying who is the author for copyright or who is the inventor for patents. The traditional interpretation is that does require a human being to fit all those categories. And that's just one of a myriad of issues that the public is facing right now with regard to AI innovation.
Matt Johnson:
Some intelligent regulation is really critical here in the high-tech space because order and predictability is important to companies. That's how they can convince folks to invest in the R&D and the technology that they're developing. If you don't have a system where companies can be confident that they're going to be able to protect their ideas through those regulations, they're not going to spend that money. And the technology development is going to slow down.
Emily Tait:
All of this uncertainty surrounding AI-related innovations, whether it's AI-related patents, copyrights, trade secrets, and of course, data, which is so important. There's so much uncertainty surrounding that. So I think that for IP clients, it's a question of we're developing our IP in-house. How do we protect it? And how do we then enforce it if it's taken away? And if we're not developing it in-house and we're looking to acquire technologies, how do we ensure that that technology is indeed protected by the company that we're acquiring or the technology that we're acquiring? So there's a lot of uncertainty right now. And I think that affects almost all clients in the space.
Matt Johnson:
Where the government might be more willing to not apply copyright protection to non-human created works, I think they realize that there is a very legitimate business interest in protecting technical innovation in the patent space. So I think they're going to try very hard to come up with paradigms that work for protecting innovation, one to encourage companies to disclose the innovations that they come up with so that others can build upon them, but really to justify all the R&D money that goes into developing those ideas so that it continues to push the technology and economy forward.
Emily Tait:
Certainly clients should be paying attention to the discussion. I mean, clients who are involved where IP is core to their business in one way or another, these developments from the USPTO this spring, there should be some interesting guidance on this point. And so I think just monitoring those developments, being aware of them and being aware of sort of the dialogue is going to be really important. And then internally just having a strategy in place, a strategy for protecting key IP is essential. A strategy for implementing AI in an ethical manner is really important.
