Frequently Asked Questions: AI and IP Policy

Basics

There is no universal definition of artificial intelligence (AI). AI is generally considered to be a discipline of computer science that is aimed at developing machines and systems that can carry out tasks considered to require human intelligence. Machine learning and deep learning are two subsets of AI. In recent years, with the development of new neural networks techniques and hardware, AI is usually perceived as a synonym for “deep supervised machine learning”.

Machine learning uses examples of input and expected output (so called “structured data” or “training data”), in order to continually improve and make decisions without being programmed how to do so in a step-by-step sequence of instructions. This approach mimics actual biological cognition: a child learns to recognize objects (such as cups) from examples of the same objects (such as various kinds of cups). Today application of machine learning are widespread including email spam filtering, machine translation, voice, text and image recognition.

Deep learning has evolved from machine learning. Deep learning uses a plurality of AI algorithms (so called “artificial neural networks”) to recognize patterns, hence being able to group and classify unlabeled data.

AI systems are viewed primarily as learning systems; that is, machines that can become better at a task typically performed by humans with limited or no human intervention.

“Narrow AI” refers to techniques and applications designed to carry out singular or limited tasks. This is distinguished from “general AI” or “broad AI”, which refer to AI systems that are able to successfully perform any intellectual task that could be undertaken by the human brain or the hypothetical ability of a machine to far surpass the human brain.

AI and related areas

Recently, AI systems have matured enough to take on tasks previously performed by and associated with humans only.

Examples of AI generated works are “The Next Rembrandt” project where AI produces brand-new paintings replicating the artist’s subject matter and style, or “Portrait of Edmond Belamy”, one of a group of portraits of the fictional Belamy family created by AI. Emily Howell or Bot Dylan, computer programs that compose music, or an AI-written novel “The Day A Computer Writes A Novel” are another examples.

AI generated output that became the basis for two patent applications, are examples of AI generated inventions. One application was for a new type of beverage container based on fractal geometry and the other was for a device for attracting enhanced attention, possibly helpful in search and rescue operations.

The WIPO Technology Trends 2019: Artificial Intelligence report provides data and analysis that identify trends, key players and geographical spread of AI-related patents and scientific publications. The data for the figures in the report can be found on the report page. Furthermore, the PATENTSCOPE Artificial Intelligence Index allows you to search in PATENTSCOPE for AI patents using an equivalent search model used in the WIPO Technology Trends Report: Artificial Intelligence.

AI has a significant impact on the creation, production and distribution of economic and cultural goods and services. AI is increasingly driving important developments across all fields and industries. Autonomous vehicles, advanced manufacturing processes and medical diagnostic tools are frequently referred to examples. It is becoming clear that AI is going to impact almost all areas. As the development of AI speeds up, its impact and general use will increase having significant impact on society and the economy. AI will start to perform many routine tasks that, until now, have been done by humans.

The WIPO Technology Trends 2019 - Artificial Intelligence reportPDF, title goes here shows AI related inventions shifting from theory to practical application.

There are a number of factors that have contributed to the acceleration of developments in the field of AI. One factor is continuously increasing computing power. Another factor is the availability of clean, structured training data (usually produced by human, hence the name “supervised machine learning”). Training data connect an input with a labelled answer or output providing a “textbook” that allows the AI system to learn and improve carrying out an assigned task.

While data have been available for a long time, it has been only in recent years that labelled, usable data have been produced in increasingly abundant quantities, for a vast range of purposes, and by a multiplicity of devices and activities commonly used or undertaken throughout the whole fabric of contemporary society and the economy. Data are an essential element in the operation of AI and are, thus, potentially economically valuable.

The fundamental goal of the IP system is to encourage innovation through new technologies and creative works. This includes human created as well as AI created inventions and works.

AI also provides a general use technology to assist in the application, management and administration of IP systems and tools.

AI and IP policy

Here, a distinction has to be made between human created works/inventions and machine created works/inventions.

Qualifying human created works/inventions are protected by the existing IP frameworks, including patents, copyright, industrial designs, and trade secrets.

There is an ongoing debate whether those frameworks and systems need to be modified for machine created inventions/works. In broad terms, the discussions regarding machine created inventions/works are focused around:

  • potential protection for the actual machine created work/invention itself. This tends to focus on the question whether AI be an inventor or creator within the existing IP frameworks.
  • potential protection of the AI algorithms and software.
  • potential rights concerning the underlying training data and data inputs.

There is also a debate around where the line between human creation and machine creation is to be drawn, i.e. how much or how little human input or guidance may be required to fall within one or the other.

Find out more about the WIPO discussion on IP policy and AI.

WIPO is a specialized agency of the United Nations and represents the global forum for intellectual property services, policy, information and cooperation for its Member States.

WIPO leads the development of a balanced and effective international IP system that enables innovation and creativity for the benefit of all. Within this mandate to promote invention and creativity for the economic, social and cultural development of all countries, the Member States have requested WIPO to provide a forum leading the discussion on AI and IP Policy.

The ongoing WIPO conversation on AI Policy is the first step in this process.

WIPO has and continues to develop its own IP management services and tools using AI technologies, creating best-in-class applications of AI for the international IP system. Find out more about WIPO AI tools.

More questions?

If you couldn't find an answer to your question on this page or through the AI homepage, then feel free to contact us.

Disclaimer: The questions and answers provided on this page serve a purely informative purpose and are not a legal point of reference. They do not necessarily represent the official position of WIPO or its member states.