INTERVIEW: Intellectual property, business and artificial intelligence

April 26 marks World Intellectual Property Day, which was established by the World Intellectual Property Organization in 2000. To mark the occasion, we speak with 3Soft’s lawyer Jakub Wyczik about whether businesses should be concerned about intellectual property law in light of the numerous solutions using artificial intelligence.

Jakub, much is being said about intellectual property rights and copyright in the context of the development of solutions based on artificial intelligence. Could you tell us first what concepts we will be talking about and what are the principal differences between them?

Intellectual property rights protect various creations of the human intellect, such as literary, artistic and scientific works, artistic performances, inventions in all industries, trademarks, industrial designs or trade secrets. In simple terms, the purpose of intellectual property law is to seek to ensure that people can benefit from whatever they invent or create. It promotes the development of a society driven by human creativity and innovation, while maintaining the proper balance between the interests of creators and users.

The foundation of intellectual property is copyright, which protects manifestations of human creative activity called works.

However, it is important to remember that the basic principle of copyright is the so-called “idea-expression dichotomy”, which is the distinction between unprotected ideas and the way they are expressed. This means that an abstract idea itself is not protected by copyright, but the way in which it is expressed may already be protected.

If we are talking about the IT market then there is no question of copyright infringement just because there are many programs serving the same purpose, e.g. word processors. However, the specific code of the program, its documentation or graphical interface may already be protected. Copyright, however, does not protect the broadly understood function or operation of software, which can be protected independently by patents granted for particular technologies.

Today we often hear about new threats to intellectual property rights from solutions using artificial intelligence. What does this look like in practice, especially from the perspective of the business world?

Much has been said about the challenges of developing generative AI, that is, AI that creates new content such as images, video, audio or text. Meanwhile, intellectual property in the context of AI is seen much more broadly. For example, when we talk about business, let’s look at the first step involved in the development of any information system: access to data.

Access to a large, high-quality data set is crucial to the successful implementation of AI in an enterprise. Artificial intelligence algorithms learn from data, so the more and better quality data we provide them with, the more effective AI systems will be. High-quality data allows for better analysis, prediction and pattern recognition, which translates into better business performance and competitive advantage. At the same time, the data must be properly organized and structured so that algorithms can process it efficiently and draw valuable conclusions from it. Without access to the right data, AI implementation may be less effective or simply impossible.

Is there any way entrepreneurs can strengthen the protection of their data?

Proper segregation of data not only allows the implementation of new solutions, but can also create new assets and contribute to increasing the protection of existing ones. What I mean here is not just about trade secrets, but also database rights. Directive 96/9/EC guarantees the protection of collections of independent data arranged in a systematic or methodical manner.

It must be proved that obtaining, verifying or presenting the contents of such a database has required a substantial investment from a qualitative or quantitative point of view. Such an investment can therefore include solutions that focus on systematically or methodically organizing the elements in the database and providing individual access to them.

Using the services of a professional provider in this area, such as 3Soft, undoubtedly bolsters the argument for the emergence of such protection. As a result we get a company asset that provides the exclusive right to perform all acts of appropriation and exploitation of the results of the investment made.

You have mentioned the results and the exclusive right to them. But doesn’t the fact that we use non-self-made solutions to which we don’t have full rights change things?

Of course, this has important implications. Currently, it is often said that AI results cannot be protected by copyright, but this in no way affects the protection of trade secrets and databases which I mentioned before. However, a case-by-case analysis is crucial, primarily in terms of the content of the contracts to be concluded.

Today, most advanced artificial intelligence systems, including those aimed at business, are offered in a SaaS (Software as a Service) model. Many people may be afraid to use such solutions not only for technical reasons, but also because of legal issues caused by unfamiliarity with the practices of contracts for such services. Meanwhile, such solutions should not necessarily always be feared. Contracts for such solutions available on the market usually explicitly state that everything the user generates remains his or her property. The system provider does not seek to have the right to the results of AI, just as Microsoft does not claim rights to all documents produced using Microsoft Office.

Many lawyers who are used to custom software contracts also expect the SaaS provider to grant the buyer a copyright on the software. Meanwhile, one doesn’t even need a license to use SaaS services, since it is done in a manner similar to web browsing. We don’t tend to worry that, as we browse the Internet, someone will sue us for using their intellectual property. That’s because we are protected by statutory exceptions derived from European laws, which allow the legitimate recipient of a service to do what is necessary to use it for its intended purpose. I could go on for a long time about SaaS contracts and the legal awareness associated with them, so perhaps this is a good topic for a separate conversation.

Jakub Wyczik – a PhD candidate at the Doctoral School of the University of Silesia in Katowice and a member of Cyber Science. He is a lawyer and researcher of legal problems in IT and Gamedev, specializing in digital goods, artificial intelligence, IT contracts and software intellectual property. He is the author of several publications on intellectual property in the life cycle of artificial intelligence, the legal status of crypto-assets and virtual goods in digital worlds. His research focuses on the concept of data ownership and trying to highlight the important difference between information and data. Jakub also popularizes science by explaining in plain language the technical details of modern IT solutions, especially on his LinkedIn profile and as a guest on various blogs. Since 2019, he has worked as a lawyer at 3Soft.