Artificial intelligence
Why do we consider Artificial Intelligence (AI) as an individual type of IP? We will explain this and much more related issues in this article.
AI is definitely not a new thing in the world. You know it from books, stories, movies, series and so on. The thing which changed in the past several years is that the AI is starting to be a real thing in our everyday lives. Basic technologies which are used in our daily lives are using AI including social media, self-driving and parking of vehicles, internet search engines, email spam filtering, transportation systems or chatbots in eshops
There are plenty of other processes in which AI is involved such as virtual reality, image processing, banking security, art creation, medical diagnosis, education, finance and audit systems. AI creating a new AI is also confirmed.
These are the reasons why huge economies are investing more and more finances into development and inovation in the field of AI. Predictions also saying that by the year 2030 the AI will be used by 70% of all the businesses around the world. The global market with AI will cross 38,8 billion USD by the end of 2025, according to actual presumptions. Meaning that AI will be giving a considerable advantage in all types of businesses.
Real advantages of AI are that it can achieve the rationality and so it can choose the best option to achieve the stated goal. To achieve this, it uses for instance analysis of huge amount of data and interaction with environment in very short time frame. AI can also learn from its previous mistakes and thus not doing same mistakes again.
What is Artificial intelligence?
When talking about AI most people imagine a kind of robot who autonomously talk, think and do a things similarly to how humans do them and all that with only a few restrictions which are „put“ into AIs code. This is not the case today.
The AI from the current perspective is the sophisticated technological tool with various level of autonomy created by human. Humans decide about restrictions and regulations which affect the AIs performance and acting. So AI is currently the computer code – software created and adjusted by humans.
Actual AI legal protection
Nowdays, AI is considered to be a software, in most countries. Thus the basic legal protection in this matter would appear to be copyright which in majority of countries does not require any sort of official registration in order to be valid. Other types of IP protection are also available.
Copyright to the work, in this case to the computer program that is AI, arises in the moment in which the work is created. By the creation of work we mean the moment in which the work is put in a fixed form. In this case it is digital form as the code is saved in some sort of digital file. Some countries such as U.S., Canada and U.K. offer the option to register the copyright which is suitable way to strengthen your rights even more. Registration of copyright may be very useful for instance in proving that your right existed in certain form and certain time.
Patent is traditional and probably one of the most expensive way to protect your AI. But it has its reasons. There were already hundred of patents granted in relation to AI accross the globe. So patentability is also an option but be sure to consult the patentability of your AI with professional legal counsel before doing so as there are dozens of pitfalls you can fall into when you do not prepare your patent protection strategy before trying to patent invention. Also note that not all software is patentable.
Trademark can be a solid option to protect the name of your AI (if it has a name). You probably plan to make a brand out of your AI name and thus you need to take care that someone else does not use this brand in relation to the goods or services you want to offer under this brand.
Know-how, trade secret and other types of IP protection can also be considered as an option but these are very situational and we can not say that every AI requires these in order to be „well protected“.
Future perspectives of AI
There are plenty of ethical and legal question regarding AI which have not yet been sufficiently answered including:
- liability of AI
- subjective rights of AI
- nature of AI actions
We will not speak about these issues as they are out of scope of this article. However we will touch the question of AI as IP type.
Briefly speaking, nowdays AI is considered as a software. And as such it is considered in most jurisdictions. This arises from actual perception of AI as programmed code which help people solve certain tasks. But what will happen if the AI do certain tasks on its behalf or on behalf of owner of AI, which is already done in many cases? Who is liable for wrongdoing of AI? Is AI a legal subject or object? What can it be in future? Many questions arises and we will show our opinion regarding this.
In our opinion, the AI will soon be considered as something different than natural or legal person as it does not simply have all the legal specifications of either of these. Also we think that some existing AI are already on the level in which we can say that it should no longer be considered as an object of law, but also not as its subject. A new legal category should be created for this purpose which would suit the actual situation. However legal framework for this still does not exists. But there are already plenty of discussions about how to legally resolve this issue.
The most probable scenario is that the AI will be cathegorized as electronical person which will be something different than natural or legal person or actual software recognition of AI. This concept is also recommended by the EU and is widely supported accross the globe. What will this means? We do not know yet, however recommendations are that this electronical person will have 2 cathegories. First will be certified electronical persons (certified AI) and second uncertified electronical persons (uncertified AI).
The idea behind this cathegorization is that certified AI will „grab“ the portion of liability from the owner of this certified AI. Therefore the owner of certified AI will be favoured and privileged in comparisson to owner of uncertified AI who will bear the full liability for the wrongful acts of his AI. Limmitation of liability of the owner will probably not be the only advantage of certified AI. First country which adopted resolution on adopting national system for AI certification is the Republic of Malta. EU also recommends some sort of certification of AI.
By making registration or certification of AI possible, the AI will get the new type of IP protection. This will supplement the existing ways of AI protection which we have mentioned above such as copyright, patent, trademark etc.
FAQ
No, you do not need to use © symbol notice in order to have your work protected by the copyright. However, the © symbol notice is a good way to show others that the work in question is protected by copyright and that rights are reserved by you.
There is no mandatory registration for your copyrighted work. However there are local copyright registries in some countries which can register your work for a fee and thus help you solve disputes in future better. All these have their own requirements and you should consult with pertinent registry to find out what are the specifics.
Every copyright registry support the registration of AI software.
You can contact us for the most suitable option for you or choose from our copyright registration bundles.