A new report by the World Intellectual Property Organisation (WIPO) sheds light on the rapidly growing field of Generative Artificial Intelligence (GenAI). The report, "Patent Landscape Report: Generative Artificial Intelligence," analyses patenting activity and scientific publications to understand current technology development, application areas, and key players.
Surge in GenAI Innovation
The report highlights a significant increase in GenAI patenting activity. The number of patent families in GenAI has grown exponentially, from just over 700 in 2014 to over 14,000 in 2023. Notably, over 25% of these filings were submitted in the last year alone, reflecting the growing interest and rapid advancements in GenAI technology.
China Leads the Way
The report identifies China as a leading player in GenAI innovation, with inventors there responsible for over 70% of all GenAI patents filed globally. This dominance is due to factors such as increased government funding for AI research and a thriving domestic technology sector. However, the report also identifies the United States (US), South Korea, Japan, and India as demonstrating significant growth in GenAI patenting activity.
Where are the most GenAI technologies invented*?
1. China: 38,210
2. United States: 6,276
3. Republic of Korea: 4,155
4. Japan: 3,409
5. India: 1,350
6. United Kingdom: 714
7. Germany: 708
*Based on numbers of published patent families between 2014 and 2023
Focus on Specific GenAI Models and Data Requirements
The report analyses the types of GenAI technologies being patented, along with the data they leverage. Among the most prominent are Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and decoder-based Large Language Models (LLMs). Notably, GANs account for the most patents (around 9,700 patent families) within these categories and primarily use image and video data.
In contrast, LLMs, which have gained significant interest due to their application in chatbots like ChatGPT, rely heavily on text and speech data. This highlights the crucial role of understanding data types in GenAI inventions. Understanding the data used is essential for IP professionals to assess patentability and potential infringement issues.
GenAI Applications Span Industries
The report identifies a wide range of industries expected to be impacted by GenAI. Key application areas include software development, life sciences, document management, business solutions, manufacturing, transportation, security, and telecommunications. For instance, GenAI has the potential to revolutionise drug discovery, product design, and customer service interactions.
Looking Ahead: A Complex Landscape
The WIPO report on Generative AI reveals a booming field with intense competition, particularly from China. Inventors and companies filing GenAI patents should act swiftly to secure intellectual property rights. Aligning your invention with a specific industry need, such as drug discovery in life sciences or product design in manufacturing, can strengthen your patent and commercial potential.
Remember, the type of data your GenAI model uses, like image data for GANs, is crucial for patentability. Considering these factors and staying updated on emerging GenAI models like diffusion models and LLMs will empower you to navigate the patent landscape strategically and maximise the impact of your GenAI innovation.
Read the full report here: Patent Landscape Report: Generative Artificial Intelligence. (wipo.int)