AI
Image Source - https://rb.gy/ofyuek

This article is authored by Devina Poonia from Ansal University, Gurugram, and edited by Gitika Jain. It explains the role of Artificial Intelligence and Intellectual Property Rights in shaping our future world and discusses the advancement of AI solutions. 

Introduction

The importance of artificial intelligence was realized during World War II when the first modern computers began to decode massive breaking machines during the war. The famous John McCarthy in the year 1956 realized the need for artificially intelligent machines which could think and analyze the situation before humans could interpret and find a solution to the problem. IP strategists used to go with the traditional methods of searching while obtaining a patent but the evolution of AI changed the whole scenario and now IP innovators and strategists felt the need to protect AI in obtaining patents. Now businesses employ AI techniques into their work environment to radically improve IP services, from neutral patent translation to AI trademark search and more. 

What is AI

Artificial intelligence is a machine that interprets signals via input received by programmed data and responds in the desired way as output and is capable of carrying out specific tasks in sync with machine learning without being programmed daily and having being fed information. Humans need not interfere in the process as their work and tasks are carried out by the machine. Machine learning is a broad variant of artificial intelligence in which they both respond in the desired output in sync with each other. Machine learning is the backbone of AI and is fed with huge amounts of data which later on carries the specific task when prompted. An example of machine learning could be scanning a document or captioning a photograph or it could be interpreting different languages. 

Download Now

Three types of AI 

There are broadly 3 types of artificial intelligence bifurcated to understand the complex relation of AI with a high level of intelligence:

  • Narrow AI: This term broadly signifies the functioning of only single and specified tasks to carry out the purpose for fulfilling and achieving limited goals. It is functioned to perform one task at a time rather than a group of tasks and is more practical in nature. Narrow AI is a more common machine learning used in day to day lives by people and is adopted by modern society. It is generally termed as weak AI as it caters to only a particular field of intelligence. Like using Apple Siri or Google maps navigating routes or a recommended list of songs on Spotify. This generation has already grasped the concept of narrow AI.
  • General AI: This term is “where we are going”. General AI has yet not been achieved by modern society but this type of AI is aimed to be implemented in the near future. It promises to tackle and process advanced solutions to problems but it can be pricey to be protected. General AI is aimed to be such an optimized solution that it can even outperform humans and their skillset but this in change requires a set of data to train them which can even lack the common sense of basic functions because of the highly skilled data to be fed in the algorithms. General AI in the near future can become competitive to the human level and can even replace the need of humans in many skillset industries as it has the qualities to learn and gain many skills which humans possess. Ray Kurzweil, Director of Google engineering, predicts that in 2029, artificial intelligence will reach the human level of intelligence. 
  • Super AI: Compared to what we discussed above about the positive elements of AI which overpowers the abilities of humans in the future, the super AI is a stronger element than the above-mentioned points. We have heard from a lot of researchers and scientists that a time will arrive in the future where robots will enslave humans or it will render humans unemployed as AI will be so advanced that it will replace mankind with machines. These robots are supposed to excel in every field namely maths, medicine, science, hobbies, sports and many more although it’s impossible to achieve this future as it will require huge investments by the countries only if they agree to adopt such technology. 

Importance of AI in IP 

As the world is evolving and the need to depend on technology is increasing, experts have noticed the ever-rising importance of artificial intelligence in the IP sector. Many years back, professionals did not understand the need for quality data and the genuineness of the fact that how it could drastically establish and lay the foundation for the growing IP sector. Now, the importance of quality data has been recognized by the professionals in the industry which eliminates the mistakes, inaccuracies, inefficient data and processes the data in a cleansed manner. The main underlying principle behind quality data is quality solutions to IP problems as a business requires high-quality data on which all the biggest decisions can be made. The integration of artificial intelligence has removed the traditional practices of patent search methods that were previously adopted for obtaining patents. The shift is more towards AI solutions as it frees up the time of professionals to focus more on analysing the better formulation of effective strategies and insights into the patent search. IP laws ensure that it focuses on novel inventions related to AI over which patents can be obtained in different countries on the basis of novelty, the capability of industrial protection, and is not too obvious.  

The growing demand for obtaining trademarks, copyrights, or patents was becoming difficult with the advent of AI, but blockchain ensures the increasing challenges faced by the IP offices. Blockchain is a set of blocks and chains where each and every block contains a piece of information and provides for verified chains and if one block of information needs to be changed it must acquire the consent of every authorized party. It can help perform a number of functions in IP like record keeping, control and track of distributors and users, evidence of rights, establishing patents or IP contracts. The real benefit of blockchain in patents is that it can be used to patent an invention in many countries through the presence of smart contracts.

Challenges faced by AI in the IP sector 

When inventions relating to AI in terms of patents, copyrights, and trademarks posed a challenge to the IP sector, various possible solutions came to the surface to tackle the problem at hand. It did not prove to be effective as AI could not tackle huge amounts of data and was not verifiable at the ends of the authorized parties. Technical issues in AI raises a lot of questions regarding ownership issues or contractual issues. We are going to discuss this in detail in the article.

  • Issues in contracts: Earlier AI used bespoke hardware systems that imitated the exact functions of a human brain but now software graphic units are implemented which saw an increase in the use of central processing units. Therefore, IP related issues arose outside the domain where software was used other than the above-mentioned elements. Challenges faced by commercial agreements arise when there is no clause in the contracts mentioning the latest developing software including ownership and licensing issues. Agreements need to mention all the vital clauses that involve new IP development software or indemnity relating to third-party authorization.
  • Collaborative model: Many AI developers face issues when third-party organizations require relevant training data in order to establish themselves in the market or obtain relevant rights over the solution. Obtaining IP rights over the advanced solutions ensures that the ownership and licensing will become of utmost importance and it will abstain the third party to involve where AI developers pay for the license to obtain training data and IP rights will ensure that the AI developers tune in with collaborators to protect their software interventions.
  • Customer data: The seller provides relevant training data authorization to customers who require the support of these training databases to work in sync with the seller’s software to adapt to customers business services. Problems arise when there is an intrusion in the cybersecurity system of the customer’s current software provided by the seller which often brings in the question regarding ownership or copyrights. Again a contractual challenge will come forward if the customers want to re-sell that software to another service provider. If IP rights have been obtained over the software, then it won’t be a challenge to protect the same but if not, then it becomes difficult for the sellers to protect their inventions in AI. Although with time it is difficult to protect the invention regarding AI software and the number of retailers is generally small who have obtained and secured their rights over their software inventions. 
  • Ownership: As AI now is able to generate 3D inventions, graphic printing, poetry, and artwork, people questioned the authenticity of the work produced by AI and elaborated on the fact that the AI also needs to be protected under IP. Human inventions have already been protected under the regime of IP laws so it became important to protect and secure AI inventions. Technical issues like the software inventions and ideas generated to develop training software posed a great challenge to be authenticated under IP. Issues came forward as if AI developers are granted patents for their input stage should they be rewarded with the same for the output stage? Should AI inventions fall in the public domain and if yes what will be the authenticity and the limit to which the patent will be granted? Their solutions probably lie on the fact that AI solutions make humans live better by providing solutions in every domain but it needs to attain a legal status among the society so that AI authors are granted IP protection. 
  • Liability issues: Inventions like 3D art, poetry, and artwork invokes copyrights infringement, in the same way, there are some AI technologies that analyze business strategies and investment decisions for a company which involve the use of certain AI software that might be patented by the owners but the service provider would be unaware of the fact as AI patents are recognized in the public domain which might bring the issue of infringement of patents in the AI world. The question which arises is, who is liable? First, to answer that, it is important to know the source of originality from where the patented technology has been imitated from, and second, can the care-takers and the in-charge authority of the AI also be held liable in cases of infringement. It is important to establish the liability of the authorities so that infringement cases can be dealt with. Also, the owner, who did not obtain patents over their inventions, can also land in trouble because there is a need to establish the said invention in the legal domain. 
  • Legislation: In order for patented AI inventions to be recognized in the legal footing, it is essential that IP laws are updated from time to time, there are huge changes witnessed by the IP sector and the challenges that new inventions and their owners tag along with, the circumstances which change the outlook of the industry and ask for new reforms so that real owners can patent or copyright their inventions. If lacunae between the AI and IP continues to exist, there would be no balance between the AI inventions and IP laws. There is a continuous need for establishing forums that single-handedly deal with AI and IP disputes. 

The main challenge faced by the IP regime is whether AI robots can be termed as inventors? A humanoid developed by Hanson Robotics in Hong Kong named Sophia gained its first citizenship in Saudi Arabia and upon being interviewed said she would like to carry on her generation. The question that arises is, can Sophia’s progeny be patented? The European patents offices have strict rules for patents that the invention would be granted a patent only if the in-depth research over its “non-technical” and “technical” features lead to an “inventive” step. Although these standards are different in many countries which poses another challenge that it may be difficult to obtain patents over such technology. 

AI intersecting the domain of IP

Artificial intelligence is everywhere. It is now inevitable for humans to live without technology in every field. There is a need for AI be it science or sports it has become essential in everybody’s lives. AI is a helpful technology in IP. It is being adopted by various industries and institutions where people thrive on advanced software solutions. IP helps in regulating and protecting AI inventions. WIPO has ensured that IP laws start recognizing the need to deploy and develop AI methods into every organization. Both AI and IP involves the use of technology-based inventions that are administered by large corporations to conduct business effectively.

IP laws can only protect the advancements and legality issues in the AI and have a significant positive impact on the economy, creation and distribution of goods and services. AI and IP intersect because the main goals of the IP policy are to stimulate creativity and invention into technological areas, both deals with the inventions which involve the use of patents, trademarks, and copyrights. AI software developers attract the need to obtain patents over their technology which again intersects IP as patents are an intellectual right and need to be protected.

AI is used in every sphere of life and it is not possible that it does not attract the IP laws, in fact now every development in the domain of AI is subject to the clause of IP and brings in the effective way to deal with the present-day issues. One such big example of an intersection in the field of medicine. The need to patent AI machines in the field of medicine became important as strategic improvements led to a breakthrough in skin cancer and experts argued to reform new measures for technology whether the technology used is patent-eligible or not. For this, the authorities laid down a two-test policy to check the subject eligibility of patents. The first test was to determine if the claims are directed to be patent-ineligible in terms of nature, abstract ideas, and natural phenomena. The second test was to determine the claims which involve a combination of elements that transform into a patent-eligible system. These two tests have ensured that medical AI inventions can be granted a patent over their work which brings in changes in the medical regime. 

WIPO, through its conferences and policies, has always highlighted the importance of AI and how it should be protected by the IP sector and must be incorporated into the business models so that issues like infringements in terms of patents or copyrights do not arise. Machine learning in the form of electronic data relies on information which is the centre for intellectual property and innovation in a global digital economy. IP laws have to be in sync with AI and have to keep closely monitoring the advancements to maintain transparency in the laws and to provide legal solutions to the AI inventions. 

AI and IP in India

It is important for India to step its game and recognize the importance of AI inventions and how it should reform the current IP laws to be in contact with the technological inventions in this domain. India is a new hotspot for currently filing patents for every new innovation. Industries need to deploy the new algorithmic data provided by the AI to make smart decisions and business strategies. In a recent report by WIPO, it showed that India is now one of the top 10 countries to file for patent inventions in AI, with over 20,195 applications filed in the year 2016. IBM has remained the top distributor in the USA charts for filing over 9,100 patents in 2018, the investors are from India and understood the need of these advanced technologies and soon became the beneficiaries particularly filing a patent for cloud computing and blockchain. It has remained the 26th consecutive year for filing patents in the USA markets. For India to be patenting AI-backed technology, one has to follow the guidelines laid down by Computer-Related Inventions (CRI) which excludes the overall patenting of algorithms of computer and computer programme. To gain IP protection for AI software, the following needs to be noted:

  • Describing the component hardware along with AI algorithms while filing for a patent.
  • Claims describing working methods in AI software.
  • Refrain from filing directly for programming codes in AI algorithms.

To get a patent registered in India is very difficult for AI technology, in the analysis below we will discuss two AI technologies that have been granted a patent in India. 

  • IN239913: This application contained claims, A proactive user for a computational device having an operating system, the proactive user interface comprising: 

(a) a unit of interface for helping in communication between a user of the proactive user interface and said operating system and interface unit including an emotional agent for communicating with the user; 

(b) at least one software application controlled by the operating system; 

The word “Artificial Intelligence”, as can be seen in claims of the two granted patents but that word, one may notice, is just used to represent part, that utilizes data/commands provided by the AI system, of the said system. No focus, however, is made on the operating principle of AI.

Is AI an asset 

Artificial intelligence is a boon for all the industries. It is ubiquitous and is present everywhere whether as Amazon’s Alexa or technology used in touch sensitivity. AI in the medical research and procedures, as well as business complexities, has proved that without AI, it is now difficult to thrive and be ahead of the competitors. It improves customer management and mines data for thousands of professionals to streamline their strategies. In the IP sector, AI is an asset for patents, copyrights, and trademarks and we will discuss how it is an asset for these IP rights. 

Patents 

AI has made major breakthroughs like treating skin cancer by using Google-algorithms and in 2016, Google developed its own Neural Machine Transmission to interpret different languages by developing their own technology. This calls for patents so that the real owner’s invention does not get recognized falsely and some other individual claims over this invention. This is where IP comes into play where a patent would be granted to the human inventor of this technology and not the AI system as the inventor. WIPO’s definition of “inventing mind” has been an arguable topic, to decide upon whether the word “mind” is human or AI. Moreover, patents and AI are beneficial for the professionals and must be patented in order to protect it from future infringements. Therefore, AI in patents is somewhat an asset to the industry.

Copyrights 

The domain of copyright and AI faces several struggles over the right of ownership and infringement so AI would be more of a bane than a boon in this domain. AI has been able to create art but not establish it’s ownership because again an inventor must be human, a machine cannot obtain copyright over their literary work. AI-artists won’t be able to rely on the principle of fair use if their work is substantially similar to the work of another AI-artist who sells and displays it. This was well explained in the case of Graham v. Prince. The plaintiff, Donald Graham, brought an action in federal district court against defendants Richard Prince, Gagosian Gallery, Inc, and Lawrence Gagosian for copyright infringement arising out of failure of Prince to seek permission of Graham to use one of his photographs to create the “appropriation art” for which Prince was well known. Prince used Graham’s photograph, entitled Rastafarian Smoking a Joint, to create an artwork known as Untitled (Portrait) (“Untitled“). Gallery was Prince’s primary gallery and agent according to the complaint.

The Court denied the motion to dismiss and alternative motion for summary judgment. It was noted by the court that fair use was not established under 17 U.S.C.S. § 107(3) because Untitled reproduced a photographic portrait without significant aesthetic alterations. It was put on  exhibition and sold at a commercial art gallery, using the entirety of Graham’s photograph. The important notes to take away from this case is that the AI-artists should not only document the creative process when selecting and inputting the underlying art, they should also consider to evaluate to determine whether the resulting AI-work is sufficiently transformative before releasing it to the public to mitigate any potential claims of infringement.

Trade secrets

Trade secrets are secrets that are protected by the investors so that they do not get leaked. It contained sensitive business information regarding a brand or design, or logos that provided for competitive advantage as it wasn’t known to others. A trade secret is often protected by law and monitored closely so as to maintain secrecy but AI developers thrive only when this information is shared because AI cannot shape its fundamental process of algorithms only form one aspect available to them, they need to be provided with this information where they can make large discoveries and develop software which can be assisted in R&D technologies. AI can be an asset in trade secrets only if relevant secrets and information is shared with the developers which helps them to innovate and invent. 

Conclusion 

AI now has proved to solve daily business complications by providing complex solutions. It is a technology that has been prevalent for decades. It can manipulate huge amounts of data quickly and effectively analyzing the most optimizing solution. Day-to-day IP management tasks proved to be a burden where analysts used to spend hours and days to conduct a relevant search for patents, thanks to AI advanced solutions where strategists don’t have to worry about finding a competitive analysis for patents. But with the growing rate of AI, there comes a point when it becomes difficult for IP portfolios to manage such huge databases and it becomes challenging for the individuals to bridge the gap between technology and protection. Over the time period, the IP sector has recognized the difficulties and have modified their policies according to the AI inventions so that it can find a place within this regime. IP professionals have the real opportunity to incorporate AI and get insights from the same which can now be easily accessed to contain an impeccable amount of data. This can, in turn, shape future investment decisions in research and development, help companies understand their relative strengths and weaknesses in terms of their competitors and enable companies to understand more about the potential opportunities in new markets. Generation of business insight, by the help of IP professionals,  that can open  new markets, accurately value an IP portfolio and deliver a better understanding of what and where the next generation of IP investment should come from, can be done efficiently.

References 

LawSikho has created a telegram group for exchanging legal knowledge, referrals and various opportunities. You can click on this link and join:

Follow us on Instagram and subscribe to our YouTube channel for more amazing legal content.

RELATED ARTICLESMORE FROM AUTHOR

LEAVE A REPLY

Please enter your comment!
Please enter your name here