Artificial Intelligence on Patent Law: Artificial intelligence (AI) is the ability of computer systems to perform tasks that require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages. Here that intelligence is demonstrated by a machine, AI realizes/understands its environment and is capable of taking actions to maximize its chance of achieving the goal.
There are various implications for AI which are majorly used for the ‘Web-Search Engines’ and ‘Recommendation System’ purposes and we can have a lot of renowned examples on it such as Google, Amazon, YouTube, etc. Apart from these, understanding human speech such as Siri and Alexa, self-driving cars by Tesla, and many more.
Artificial intelligence, or the rise of computers capable of independent problem-solving and even independent creativity, is the most complicated and powerful danger to the intellectual property regime that has ever existed. Until recently, AI lacked a clear and compelling application in which it could play a significant role.
Yet, now it has been enlisted in an increasingly varied range of societal tasks, its impact on future societal trends is being aggressively disputed. Finding the proper answers to these concerns will necessitate the participation and involvement of all key parties from around the world.
To guarantee that the most advantageous revisions are implemented, considerably more education on how the philosophical position on IP is established and maintained as required, so that all stakeholders may push for a result that represents their interests. The evolution of AI is being fuelled by increasing stocks of large data and breakthroughs in affordable high processing power. The invention, production, and distribution of commercial and cultural goods and services are all impacted by AI.
“Artificial intelligence is set to radically alter how we work and live, with great potential to help us solve common global challenges, but it is also prompting policy questions and challenges.” AI overlaps with IP in a variety of ways, as one of the key goals of the IP system is to foster innovation and creativity in the economic and cultural sectors.
So, before we assess Artificial Intelligence on Patent Law, we also need to look at the factors where IP and AI are clashing and fix them according to the needs of the time.
Artificial Intelligence and Intellectual Property
The World Intellectual Property Organization (WIPO) has begun to focus on aspects of AI that are special to IP, as policymakers they began to grasp the wide-ranging consequences of AI and how AI can bring change in the existing system of intellectual property and that changes will be positive or negative to this field. Multiple strands are running through this interaction. Now WIPO is working on how AI could be useful for intellectual property, WIPO is working a lot to go parallel to these technological innovations.
AI applications are increasingly being used in the administration of IP protection systems. WIPO Translate and WIPO Brand Image Search are two AI applications that use AI-based applications for automatic translation and image recognition. Other AI applications have been created and deployed by several IP offices throughout the world.
WIPO held a meeting in May 2018 to address AI applications and to encourage the exchange of information and sharing of such applications. WIPO intended to compile a list of the most pressing challenges and concerns about IP regulation that have arisen as a result of AI’s growing use as a general-purpose technology.
In September 2019, a Conversation with the Member States and representatives from the commercial, research, and non-governmental sectors was held at WIPO for this purpose. WIPO is in close contact with the agencies in charge of these disciplines, and the WIPO Conversation on IP and AI will contribute to the various strands of the ongoing talks.
WIPO, for example, is taking part in AI and digital platform roundtables organized in response to the UN Secretary-High-Level General’s Panel on Digital Cooperation in the Age of Digital Interdependence report’s recommendations. WIPO is also participating in the Road to Bern via Geneva conversations on digital and data cooperation in the run-up to the 2020 UN World Data Forum in response to this report.
Impact of Artificial Intelligence on Patent Law
Patents are well-known for being given as part of a statutory bargain: they are time-limited monopolies granted to private parties in exchange for the public disclosure of innovative innovations (and eventually, into the public domain). While the patentee receives a set of temporary exclusive rights, the technical information given by the patentee can be used by competitors and the general public when the patent expires after 20 years.
Patents are only awarded for inventions that are of relevant subject matter, original, have an innovative step, and may be used in industry. Artificial intelligence (AI) has ushered new ways of innovating that require only a minimal amount of human engagement.
This has prompted several fundamental questions, the most crucial of which is whether the patent system can still achieve its goals of innovation and protection. Artificial Intelligence on Patent Law and generated inventions may not differ significantly from other computer-implemented inventions in these areas.
However, it appears that Impact of Artificial Intelligence on Patent Law is growing, and there have been situations where an AI application has been cited as an inventor in a patent application. It is believed that the growth of AI technologies necessitates immediate changes to the patent system to avoid severe negative repercussions from an unequal level of protection for AI outputs, which could have harmful social, economic, and ethical ramifications.
Patentable Guidelines
Guidelines state that patent protection is only available for inventions that are in the fields of technologies while they are not excluded from patentability as per ‘Article 27 TRIPS’. The patentability of any software is not harmonized globally and exceptions are defined at the regional and national levels.
The best example suitable for this is Article 52 which is of the European Patent Convention (EPC) states that any patentable invention by computer programs shall not be considered. Under the EPC, it is generally understood that inventions should have a technical character which is the so-called computer-implemented inventions should receive patent protection, whereas computer programs are not accepted.
The question of whether AI-generated inventions are patentable must be considered in light of whether patents on AI-generated inventions would advance the patent law system’s core goals. Some suggest that providing patent protection to AI-generated creations will spur innovation.
Software is not specifically excluded from patentable subject matter in the USA. Therefore, in some jurisdictions, some inventions related to software or computer programs may be considered patentable, while in other jurisdictions it might not be the case.
Ownership and inventorship
We all know that Inventorship and ownership are two different things. The inventor, or the person who develops and achieves the invention, can be the patent’s owner (for example, if he/she applies themselves).
In practice, a legal entity, such as a company, files and prosecutes the application for, say, an employee’s invention, and thus becomes the patent’s owner. As a result, the inventor (typically a company/applicant employee) is left with little more than the moral right to be called an inventor (albeit an employee may be paid a bonus or get a royalty based on contractual terms).
Indeed, under European Patent Convention (EPC) regulations, a European patent automatically belongs to the inventor or his or her successor in title; in circumstances where the inventor is an employee, whether the latter or the employer is the owner is established by national law.
The European Patent Office (EPO) noted in its DABUS judgment that computers and AI systems currently do not have any legal rights, including the right to hold a patent (and cannot be designated as inventors, either). It held, in particular, that “AI systems or machines cannot have rights that result from being an inventor” due to their lack of legal personality.
The EPO however confirms that the AI’s owner may be entitled to ownership of any qualifying output generated by the machine if certain circumstances are met. This is consistent with the position of the artificial Intelligence project team the group responsible for filing the DABUS application which has maintained that while the AI should be recognized as an inventor, the patent should be owned by the AI’s owner. (Artificial Intelligence on Patent Law)
If machines are to be accepted as inventors in the future, and an AI system (e.g., IBM’s Watson) generates an invention while being operated by a third party, attributing ownership to the end-user could hurt social welfare, given that AI developers deliberately limit access to their AI.
Rather than restricting access to the system, assigning ownership to the developers may encourage them to seek to enable it. A programmer who just generates general-purpose claim-generating code, on the other hand, does not contribute to a significant part of the claimed invention; rather, this code will be used by others (i.e., end-users) to make such contributions.
Finally, Banterle claims that allocating ownership to end-users is the best “economically efficient” alternative because they “hold these patents in the highest value and so aggregate welfare is maximized.” This is an extremely challenging problem to tackle, thus taking a case-by-case approach is likely to be the best option. However, because machines cannot be classified as inventors, this topic is currently only hypothetical.
Alternative Approaches on Artificial Intelligence on Patent Law
While the focus of this research has been on whether AI-generated innovations qualify as a patentable subject matter under the current legal framework, it’s worth emphasizing that not everyone feels that such output should be protected in the first place.
Numerous commenters suggest that AI-generated inventions should be released into the public domain without human intervention. Personality and incentive-based theories which are used to support IP rights are mainly inapplicable when it comes to such creations, according to some.
For example, Yanisky-Ravid and Liu claim that these theories were created primarily with humans in mind and that they are irrelevant in this context because AI systems are “independent, creative, unexpected, and rational”.
Similarly, McLaughlin emphasizes the contrast between computer-assisted and computer-generated inventions. In light of the “prevailing theoretical reasons” for IP rights, he believes that an invention should enter the public domain if there is a significant “nexus to human inventorship” – which would be the case where AI-produced inventions are generated autonomously. In a recent position paper, the UK Chartered Institute of Patent Attorneys (CIPA) emphasized the problem of requiring a sufficient level of human participation.
While many in CIPA believe that patent protection should be available as long as the relevant substantive provisions are met regardless of whether the invention was created by AI others believe that existing inventorship requirements should be maintained and that patent protection for AI-generated inventions should be available only to the extent that there is a “genuine human contribution”.
Conclusion
We all understand the importance and the level of utility of AI not only in day-to-day life but also for technical advancement and jurisprudence. As we know everything has its pros and cons, and AI too has its own. Currently, AI is neither in a position to acknowledge the need for replacement invention by itself nor initiate any inventive process.
Additionally, AI can only access the information that is available electronically, and currently, no AI has the potential to truly access all the sources of information available for invention. However, AI systems will keep evolving and they will increase in utility and availability, which influences the need for patentability. On the other side, AI can make the whole process of innovation smoother, quicker, and cheaper.
There will be an increase in the importance of legal systems like the supply of the latest information through the patent register or the likelihood to trade intangible goods. AI not only re-considers the applicability of several patentability issues but also the set-up and functioning of the general legal system.
In the end, we can conclude that impediments to innovation through AI are not something that is to be feared and should be welcomed. After all, if technological changes bring an influence on law, its attitude should be sharpened on how it should be shaped for social profit.
The “governance” of Artificial Intelligence on Patent Law has a significant impact on innovation, the economy, and society. If given rapid advancement to AI, all key stakeholders patent and non-patent experts alike must engage in proactive research and conversations to develop ways for the patent system to foster innovation while limiting any negative societal and ethical consequences.
This discussion is going on because the law is behind the technology, if the evolution of law matches the speed of the technology, then there will be no clashes based on technological evolutions such as AI.