Home News Intellectual Property (IP) challenges in the context of artificial intelligence (AI)

Intellectual Property (IP) challenges in the context of artificial intelligence (AI)

June 29, 2023

Artificial Iintelligence generated content can cause intellectual property infringements (trademarks, designs, patents or copyrights). It also affects closely related rights such as data protection or ICT.   

  1. Copyright: AI can be used to infringe existing intellectual property rights. For example, the use of AI to circumvent copyright protection measures or to plagiarise content can raise legal issues.  
  1. Patents: 
    -Patents are exclusive rights granted to an invention for commercial exploitation. However, it is difficult to determine whether an invention created by an AI is eligible for patenting and, if so, who should receive the patent. The key question is whether the invention is considered “non-obvious” and whether it involved a significant human creative process (inventive step).  
    – If a company uses AI technologies that are covered by another patentee’s patent without authorisation, it may cause patent infringement.  
  1. Designs: 
    -AI can be used to generate designs automatically. However, originality (singularity in Spain) is a fundamental requirement for registering a design. The challenge lies in determining whether an AI-generated design meets the originality criteria established by intellectual property laws.  
    – Originality of AI-generated designs. This involves assessing whether the AI has used sufficient human creativity or is simply replicating existing patterns.  
    – AI design infringement: AI can also be used to infringe registered designs. AI algorithms can scan and analyse existing designs and then generate copies or variants that infringe the design rights of third parties.  
  1. Trademarks:  
    -Trademark generation: some companies are using AI to generate new trademarks. It is important to ensure that the generated trademarks do not infringe existing trademark rights and comply with legal requirements.  
    – AI is used to personalise advertising and product or service recommendations based on consumer profiles. However, it is important to ensure that AI does not promote counterfeit products or infringe the trademarks of third parties by displaying misleading advertisements or recommendations.  
    – Use of trademarks in domain names: If an AI is used to generate automated domain names that include trademarks without authorisation, this could infringe trademark rights. Creating websites with domain names that use trademarks without consent may be considered trademark infringement.  
  1. Ownership of models and algorithms: AI models and algorithms developed have significant commercial value. Determining who is the rightful owner of these models and algorithms can be complicated, especially in situations where several persons or entities have contributed to their development.  
  1. Data protection and privacy: AI relies on access to large amounts of data to learn and improve. However, the use of data can raise issues in terms of privacy and the rights of data owners. Improper collection and use of data without proper consent can lead to legal disputes and privacy violations.  

(The above paragraphs have been provided by the ChatGPT, after consultation on various questions about IP challenges in AI).  

From the above, we can deduce that AI itself has been configured to be “aware” of the challenges posed by the use of this technology in IP.   

Indeed, there are a multitude of controversies and/or potential intellectual property infringements with the use of artificial intelligence that we will discuss below.   

Undoubtedly, we are seeing the emergence of competing interests between Authors/IP right holders, AI companies and the general public/consumer.  

For example, Drake and The Weeknd’s song created by AI posed a challenge to the music industry. 

AI voice generators

The use of a copyrighted work, e.g. music created by an artist – and his voice – should require his consent to be used on an AI platform, as we cannot ignore the fact that the author has the exclusive right to reproduce, distribute, display, transform or create derivative works of his original work.   

On the contrary, his or her professional activity would be diminished, voice falsifications could be favoured, and the interests of technology would take precedence over those of human creativity.   

In the same way, clones of human voices (e.g. D-ID), although they may favour technological progress, the lack of regulation could facilitate IP infringement, or be used for fraudulent purposes.   

AI image creation

There are also AI imagers, e.g. Stable Diffusion, (recently sued in the USA, among others for IP infringement and unfair competition).  

It seems that the developers of this type of software nourished the AI platforms with millions of original works (a priori unlicensed).   

It should be noted that it is thanks to the original works that AI images can be subsequently formed. I would like to point out that the criteria for the “creation” of artificial intelligence images through the use of technology are determined by the platform users themselves, for example, by indicating a colour or describing the desired image in words.   

However, these parameters were already pre-established by the creators of the artificial intelligence platform themselves (through the use of original works), so it is interpretable that the AI images are not sufficiently transformed.   

Therefore, we would be dealing with derivative works that could involve an infringement of intellectual property rights of the original works.  

Contextualise that there is no harmonised legislation on copyright, so there could be different interpretations depending on the territory, e.g. USA or EU.  

In my opinion, Intellectual Property rights should protect the value of human intellectual creativity over other interests such as AI.  

As for trademarks, while their use could be beneficial to track or monitor counterfeits, we cannot overlook that a lack of specificity in their regulation could cause the opposite effect, e.g. the online promotion of counterfeit goods, the creation via AI of a distinctive sign identical or similar to a priority trademark, or the use of priority trademarks as a domain name, would inevitably lead to an infringement of IP rights.  

With regard to patents (and utility models), the inventor named in the application is required to be human (created by his own inventive activity).   

From another angle, the incorrect regulation of artificial intelligence could result in a user of this technology making use of a patent without a licence from the patent holder, which would result in patent infringement.   

A similar situation would occur with designs, for if the usefulness of AI in the creation of Industrial designs is not disputed, neither can we ignore the fact that the lack of regulation could lead to infringements of previously registered designs.   

While it is true that the European Parliament has recently tabled amendments on the future Artificial Intelligence Law, to address, among others, the complex IP controversies related to AI, I consider it necessary to implement it as soon as possible, in order to find a balance between IP rights and innovation (AI), and to ensure the use of this technology in a transparent and ethical manner. 

Article by Xavier Badia.

Do you want us to help you?

Contact us and we will put a team of experts at your disposal.