Today every business is a digital business and has valuable digital assets that they might not even realize. Consider about the following scenarios, for example: A textile company develops software capable of modeling the best types and combinations of fibers needed to make a desired product; a luxury jewelry company develops a blockchain-based system of reliably and accurately tracking the history of its gems, ensuring their origins are traceable; a retail business creates an AI-based approach for presenting their wares to consumers on their website; and a manufacturing business develops an alternative automated approach for controlling their equipment that leads to reduced manufacturing times or costs.
On a day-to-day basis, these businesses would undoubtedly describe themselves as being in the retail and/or manufacturing sectors. However, you can see that as they continue to develop, they have become retail and manufacturing businesses with a significant digital element. And these digital elements have value that remains untapped – from a financial and commercial point of view – if it remains hidden, ignored and/or unprotected. In order to capitalize on such value, companies need to capture and leverage it in the same way as the other intellectual property (“IP”) at the heart of a business.
What Digital Assets Do You Have?
Let’s look at the three key registrable IP rights that play a role in the retail environment: patents, designs, and trademarks. When it comes to protecting digital inventions, patents might be the first IP right that comes to mind. However, patent protection may not be available in certain situations, as it is difficult (if not impossible) to protect known algorithms or systems applied to new scenarios, the automation of known processes and data itself. The caveat here is if the inventor can show that the invention has a technical effect – or in other words, if the improvement created by the invention: (1) measurable; (3) repeatable; and (3) objective. Once it an inventor has identified the part of the invention meets these three criteria, the next question is whether the remaining elements of the invention are not obvious. If it seems that even part of the invention meets these requirements, the software could very well be patentable.
Even if an invention, such as the specific process for tracing/recording a product’s lifecycle, is not patentable, the relevant company may be able to protect such digital innovations using trade secrets. While many businesses mistakenly think they do not have any valuable trade secrets at play, the truth is, if your business has commercial value (now or in the future), has information that is kept secret, and has been limited in terms of accessibility to only need-to-know people within your business you have trade secrets. If digital IP is not patentable, trade secrets can often provide the desired protection because they can extend to multiple facets of your technology, such as a formula, a practice, a process, processed data, an algorithm, or code as well as how a particular piece of tech works and/or the way you use data.
In the United Kingdom and European Union, designs can also be hugely important when looking at how to protect digital innovation. In the retail context, a company’s website might be extremely intuitive, user-friendly, and unique. That user interface (“UI”) might produce a world-class user experience (“UX”) that required significant resources to develop. The look, feel, and functionality of such a site can be protected as a registered design right. Similarly, an app that a fashion company, for instance, has invented to engage with consumers – which may well use an easy-to-use color-gradient slider that is completely different to what is being used by their competitors – could be protected as a design. Beyond that, maybe you have developed your own icons and animations that are used on your e-commerce site and/or in your digital marketing. Again, these can be protected by either static or event animated designs and protected in a fraction of the time and at a fraction of the cost of pursuing patent protection, making designs an effective proposition in areas that innovate at a frantic pace.
Still yet, while businesses have traditionally protected their company names, logos, and even taglines with trademarks, if your business has developed an app, chances are, your consumers or clients are interacting with your app icon on a regular basis. These digital assets are more compact than your name and might be simpler than your logo, but it is that little icon that consumers often come to really associate with your brand through repeated interaction, making it a valuable indicator of source.
What About AI Inventions?
Artificial intelligence (“AI”) currently sits at the center of digital innovation. However, there are several risks that companies will need to navigate when it comes to protecting the data behind AI innovations. These include: (1) finding the best way to ensure you can use your AI for what you want to use it for (both now and in the future); (2) proving you actually own both the data and the algorithm being used to drive the AI; and (3) making sure the tech and the underlying data in your AI invention are not accessible by competitors.
It is also important to remember that serious questions surrounding the patentability of AI/machine learning-developed innovations remain unresolved around the world. Several “test case” patent applications have been filed around the world that list an AI program – called “DABUS” – as the sole inventor. The question as to whether AI can be listed as an inventor has largely been met with a firm “no” from IP offices, which has led to these applications being rejected. (The U.S. Copyright Office rejected an application for copyright registration has similarly been rejected for DABUS-created art, and a district court in the District of Columbia subsequently agreed with the Office’s determination.) However, appeals are ongoing, including in Thaler v. Perlmutter, which means that the situation is unlikely to be determined for some time.
So, what does this mean for protecting IP relating the use of AI/ML, either in the development process of non-AI related inventions, or in inventions relating to improved AI/ML techniques? While listing an AI as an inventor for a patent (or copyright) application does not appear to be a good idea at least for the moment, this does not mean that IP relating to AI/ML cannot be protected.
THE BOTTOM LINE: Care must be taken to ensure that protection can be obtained for the key innovations relating to this subject and thus, companies are encouraged to develop an effective strategy for protecting IP relating to AI/ML innovations as they closely monitor developments in this space.
Christopher Hartland is a senior associate at Potter Clarkson, with a research background in condensed matter physics and magnetic resonance.