AI has expanded rapidly in the last decade and is now immersed in almost every area of human development, from health care and law to writing and art. The international community has recognized a need to regulate this rapidly developing technology, but its far-reaching nature makes it difficult to assign any pre-existing organization to be its caretaker. For example, the Intergovernmental Panel on Climate Change might be able to opine on AI’s impacts on the environment, but it can not advise UNESCO how to use AI to protect World Heritage Sites. As a result, many of these organizations have created proprietary programs to address the way AI impacts their projects specifically. 

One example of this can be seen in the mechanisms governing intellectual property (IP). The UN’s World Intellectual Property Organization (WIPO) aims to encourage and protect human innovation. WIPO recognized the transformative capacity of AI over a decade ago and has offered an interesting path for how to develop institutional processes to address this new technology’s development. 

The Cairo Review spoke with Walid Mahmoud Abdelnasser, former director of the Division for Arab Countries at the WIPO in Geneva from 2015 to 2024, to hear his personal views and perspectives on how multilateral organizations can develop their approach to AI. 

Cairo Review: When did AI really “start”? 

Walid Mahmoud Abdelnasser: The essence of what we call “artificial intelligence” can be traced back to ancient philosophy, through science fiction in the early 19^th century, such as Mary Shelley’s famous Frankenstein (1818), and to published scientific research  starting in the 1950s, and more precisely Alan Turing’s 1950 publication, Computer Machinery and Intelligence, which proposed a test of machine intelligence called The Imitation Game. This work is considered the birth of AI. 

Similar to previous technological revolutions that changed life on earth, like the First Industrial Revolution in the 18th century, AI has brought with it challenges and opportunities. As in those previous cases, humanity will pass through periods where there will be a continued need to adjust; minimize risks and maximize benefits. In the past five years alone, digital technologies in general, including AI, have grown 172 percent faster than average; AI in particular grew by the astronomical figure of 718 percent.

That’s an impressive jump. How would you say the world has responded?

The question of AI has not been considered seriously as an item on the multilateral global agenda and, more specifically, by international organizations with a universal membership, until only slightly more than a decade ago.

Since that time, almost all countries of the world—the exception being those passing through turbulent internal conflicts or wars—became deeply interested in exploring the full potential of AI, getting to know more about it, learning how to derive and maximize benefits to serve national or collective interests. Countries were equally keen to protect themselves from any potential cross-border repercussions or risks that might be inflicted upon them by adversaries, whether at the strategic, political, economic, cultural, social or scientific/technological levels, such as cyber security attacks or similar dangers. 

Such interest expressed and demonstrated by an increasing number of countries of the world has led to some sort of race and/or competition among a number of international organizations and fora.

What type of competition?

They were debating which among them, if any, would have the credentials, competences, capabilities, resources, mandate, and qualifications to initiate and manage the elaboration, evolution, and development of a global multilateral regime handling AI. 

What would this “multilateral regime” include?

It would need to provide potential institutional structures, agreements, rules, norms, regulations, international cooperation mechanisms, and operations for AI management. The original idea was that there could be a globally agreed-upon roadmap and agenda to proceed on this front—even if just accepting the lowest common denominators shared by all countries in the world. This would imply setting up a universal global body, or making use of an existing one, to be administered by the international community at large in a multilateral framework characterized by universality, sovereign equality, and a transparent, well informed, and democratic decision-making process.

Is such an undertaking possible? 

Well, one needs to be realistic, rather than idealistic, in looking at international positions regarding such efforts to regulate AI and its future developments to ensure its positive contribution to international cooperation. The more developed countries in the area of AI initially hesitated in agreeing to a multilateral regime to supervise AI usage, so as not to be bound by any rules that might arise and hinder progress and that would be considered restrictive from their perspective. 

On the other hand, countries lagging behind on AI wanted to avoid the problems that arose when the internet became widespread, as it continued for many years without any international body to regulate it. When the international community began to develop conventions on the internet, many developing countries saw this as a rather late and incomplete process. As a result, these countries want to start the international multilateral process of managing AI use as early as possible.

However, both countries and international organizations soon realized that AI in its entirety and with all its multiple connections and interactions with almost all areas of human activity, does not fit in just one international organization’s mandate or competence; no single organization or forum has the resources and capabilities to address all dimensions of its usage. One of these dimensions is how AI relates to intellectual property. 

So since AI is involved in so many fields, it’s too big for any one multilateral organization to handle all of its different aspects. Which organizations handle AI’s impact on IP?

The United Nations-affiliated Geneva-based World Intellectual Property Organization (WIPO), is one of those international organizations which first considered the question of AI informally more than a decade ago, under the leadership of its then Director-General Dr. Francis Gurry. After quiet but lengthy consideration within the secretariat of the organization, including the involvement of member states as well as leading AI consultants, WIPO decided to confine their handling of AI to its Intellectual Property (IP)-related aspects. 

What is the connection between AI and IP?

While AI has been increasingly impacting the production and distribution of economic and cultural goods and services, one of the main objectives of the IP policy is to stimulate innovation and creativity in economic and cultural systems. WIPO’s decision to handle only the IP-related aspects of AI was both logical and normal. Pointedly, in the 2014-2023 decade there were 54,000 Generative AI patents, and those constituted 25% of patents in 2023 alone.

The decision required two measures to be undertaken by WIPO at both the intergovernmental level and the level of the secretariat.

What was the first measure the WIPO took?

First, WIPO realized that it should not rush into establishing an institutionalized and permanent intergovernmental body to handle the relationship between AI and IP, as it was too early to embark on such a path, particularly as there are still many unclear or inconclusive aspects of such a relationship. Therefore, WIPO started by calling upon member states and other stakeholders to convene what was agreed to be called the “AI Conversation”, as a rather flexible and open-ended forum that would allow increased knowledge about AI–IP relations. 

The forum brought together worldwide experts on the subject from academia, research institutions, and the private sector, sharing and exchanging experiences, best practices, and lessons learned in this area among member states and the WIPO secretariat. In doing so, there was full awareness of the existing gaps in digital infrastructure among various countries and regions of the world.

What was the second measure WIPO took?

Second, WIPO proceeded to establish a unit within its secretariat to handle this new and important subject matter to be added to its already busy and expanding mandate. Here again, the idea was to start small and gradually develop that unit and add to its staff the expertise needed. However, when naming this unit the decision was to go broader than AI, so the name of “IP and Frontier Technologies”, was chosen for that newly established unit. The idea was to cover AI, but also “big data”, “the metaverse”, and other related subjects. Originally starting with one person, this unit evolved over time until it grew into a full-fledged division whose work program was streamlined into the mainstream work program of the organization.

When did this whole process start?

The first session of this conversation was held in September 2019, with large-scale interest and high level participation from most of the member states. Due to the lockdown in Switzerland because of the COVID-19 pandemic, the second session was held virtually in July 2020. The move to virtual mode of holding meetings, then to hybrid mode, increased world interest in following the AI Conversation sessions. This led them to start holding it twice a year rather than just once annually. This has been the case since November 2020.

What are the goals of these meetings?

The objective of such idea exchange has been to provide the basis for a shared understanding of the main questions that need to be addressed in the context of the relationship between AI and IP. There has been, as expected, a particular focus on the main legal questions raised by AI as far as IP rights and IP policy are concerned, including questions of ethics, privacy, and standards. 

The focus by WIPO on the regulatory matters related to the relationship between AI and IP continued to be a main theme on the agenda of the successive sessions of their AI Conversation. Items presented and discussed included the question of AI regulation and how it relates to contractual arrangements. In this context as well, WIPO’s Secretariat managed to design an IP Policy Toolkit particularly aimed at helping regulators and IP offices in addressing IP and AI issues in a more and better structured manner.

How does WIPO handle the role of AI in business?

WIPO has continued to pay attention to the interlinkages within the AI–IP relationship on the one hand and its related business activities on the other. Such focus has helped attract stakeholder attention on the impact of AI on IP. To serve these purposes, the WIPO secretariat developed a guide on generative AI for enterprises and entrepreneurs to help in evaluating and mitigating IP risks in the adoption and deployment of AI tools. Furthermore, and since 2019, WIPO has held a series of sub-regional, regional, and inter-regional “IP Management Clinics”, whose programs were specifically tailored to address the needs of AI-driven small and medium enterprises (SMEs). One important program that started in this area was the one based on cooperation between Japan and the Arab region, holding its first session in Tokyo in October 2019, its second session virtually in 2021 and its third session at the Headquarters of the League of Arab States in Cairo in February 2023.

So AI might present a risk regarding IP that businesses need to be wary of. But is there any way that AI can help with addressing IP issues?  

Yes, there has been significant attention on the rapidly growing role of AI in IP administration because AI applications have been increasingly used in the administration of applications for IP rights protection. For example, WIPO itself had developed two AI-based applications, namely WIPO Translate for automated translation and WIPO Brand Image Search for image recognition. 

Several government-affiliated IP Offices around the world, whether at the national, sub-regional, or regional levels, have also developed and employed other AI-based applications. In this area, WIPO has been keen to promote exchange of relevant information among its member states as well as the possible sharing among these countries of the AI-based applications used in IP administration.

A big component of IP is patenting. How does AI interact with this aspect of IP?

Yes, patents are very important because they constitute the backbone of revenues generated by IP rights registration and protection. In this area, there has been a growing differentiation between AI as an inventor on the one hand versus AI-based and AI-assisted inventions on the other. 

What do you mean by “inventor” versus “assistant”? 

One issue is when the artificial intelligence itself is the inventor of the property, which raises questions and challenges about who is named the inventor in the patent. The other issue is when the inventor is using AI assistance, which raises questions about defining and determining the extent of human contribution in the invention. This leads us to questions about “joint inventors” and the “patentability” of inventions involving AI. A new creation must be inventive to be patented, so determining whether an AI-assisted project is eligible for patent is an important step. Is the AI itself the inventor, or is it just a tool that an already skilled and knowledgeable person is wielding? In this context, it is also important to discuss how to protect the IP rights of data used in the training of AI models.

These seem like really complicated questions. Is there any existing framework for these international bodies to follow when answering them?

The relevance of the AI impact on patents as far as the international agenda is concerned could be better understood if we realize that there has already been international cooperation in the area of transfer of technology through the patent system, particularly in the areas of technical assistance, human and institutional capacity building, research and development collaboration, and licensing. 

Some of this cooperation takes place through WIPO and its various organs, including through Funds-in-Trust established by some member states to foster cooperation with other individual countries or groups of countries, while other forms of cooperation takes place bilaterally or at the regional or sub-regional level, such as through the European Union (EU), the ASEAN, and the League of Arab States (LAS). 

One example of this is the impact of AI on healthcare and how it is linked to licensing rules and practices involving medical technologies. This subject comes under the mandate of three international organizations, namely WIPO, the World Health Organization and also constitutes an important segment of the famous World Trade Organization (WTO)’s Trade-Related Aspects of Intellectual Property Rights (TRIPS) Agreement. The three mentioned organizations have successfully managed to develop, for almost a decade, a trilateral coordination mechanism that also covers other areas which come under the mandates of the three of them.

One of the important and positive outcomes of the previous sessions of the WIPO’s AI Conversation has been the increased realization that as some of the issues raised go beyond the mandate of WIPO, there is a need for a coordinated approach among relevant international organizations and fora in order to be better positioned to address the broad and cross-sectoral issues raised by AI. 

How have these international bodies responded to this need?

To this end, and at the bilateral level with other international organizations, WIPO has been regularly cooperating with two other United Nations system-affiliated and Geneva-based international organizations, namely the International Telecommunications Union (ITU) on the “AI for Good” initiative, and the World Health Organization (WHO) on the “AI Health” initiative. 

WIPO has been equally supportive of the efforts undertaken by the Paris-based United Nations Education, Science and Culture Organization (UNESCO) to develop the first global normative instrument on the ethics of AI. 

At a broader level, and from the perspective of the United Nations system at large, WIPO has been collaborating with various UN agencies to engage in broader high-level discussions on AI and the Global Digital Compact.

The Global Digital Compact is a multi-stakeholder action plan on digital cooperation, originally proposed by the current UN Secretary-General António Guterres in his “Common Agenda”, which is expected to provide an inclusive global framework to overcome digital, data and innovation divides for a sustainable digital future.

Equally important, and at the level of the WIPO Secretariat, the WIPO’s AI Conversation sessions positively contributed to empowering the relevant WIPO staff. This particularly includes those working in the IP and Frontier Technologies Division and those working more generally in the IP Infrastructure and Platforms Sector. These sessions have equally helped guide those working in the “IP for Business” Division and in the more general IP and Innovation Ecosystem Sector. Additionally, they have provided important guidance on emerging issues like “IP and the Future” in the framework of the IP and Global Challenges Sector, by enabling staff to acquire more knowledge and to upgrade their technical expertise.

What does the future hold for AI and IP?

In the process—and as far as AI is related to IP—there remains the need to set rules, norms and other regulatory measures at both the domestic and international levels to enable promoting and further expanding the innovation and creativity ecosystems through AI tools. This must be done while ensuring that all countries and regions of the world can equitably benefit from such potentials and that countries lagging behind on digital transformation are equipped, through international cooperation, with the necessary knowledge, information, and expertise to move toward closing or narrowing the digital gap. 

IP appeared in the first place to encourage, promote and enhance human innovation and creativity, but AI is changing the ways and means human beings are undertaking innovation and creativity. Therefore, there is a need to ensure that human beings remain at the core and centre of the innovation and creativity ecosystems.