New York, 3 July 2024
China-based inventors are filing the highest number of generative artificial intelligence patents, a new report from the UN World Intellectual Property Organization (WIPO) shows.
According to the agency’sPatent Landscape Report, between 2014-2023, more than 38,000 GenAI patents came out of China, six times more than those filed by inventors in the United States, which came in second place.
GenAI, or Generative AI, allows users to create content including text, images, music or software code, powering a range of industrial and consumer products, including chatbots such as ChatGPT, Google Gemini or Baidu’s ERNIE.
A game changer
“GenAI has emerged as a game-changing technology with the potential to transform the way we work, live and play,”saidDaren Tang, WIPO Director General.
Since the 2017 introduction of the deep neural network architecture that is now synonymous with GenAI, the number of GenAI patents has increased by more than 800 percent through 2023,according to the WIPO Report.
The sharp rise in patenting activity reflects the recent technological advances and the potential within GenAI.
For our common benefit
The WIPO head added that by analysing patenting trends and data, the agency hoped to give everyone a better understanding of where this fast-evolving technology is being developed, and where it is headed.
“This can help policymakers shape the development of GenAI for our common benefit and to ensure that we continue to put the human being at the centre of our innovation and creative ecosystems,” he said.
Report’s key findings
According to the report, GenAI patents currently represent 6 per cent of all AI patents globally.
The top ten applicants include Tencent (2,074 inventions), Ping An Insurance (1,564), Baidu (1,234), Chinese Academy of Sciences (607), IBM (601), Alibaba Group (571), Samsung Electronics (468), Alphabet (443), ByteDance (418), and Microsoft (377).
In terms of location, China (38,210 inventions) led the field, far outpacing US (6,276 inventions), Republic of Korea (4,155), Japan (3,409) and India (1,350).
Image and video data dominated GenAI patents (17,996 inventions), followed by text (13,494) and speech or music (13,480).
GenAI patents using molecule, gene and protein-based data grew rapidly (1,494 inventions since 2014) with 78 per cent average annual growth over the past five years.
Source – WIPO/United Nations
Patent Landscape Report – Generative Artificial Intelligence (GenAI)
Generative AI is booming. It is a cutting-edge technology that is poised to disrupt various economic, social, and cultural sectors, and it extends far beyond simple human-like text generation using chatbots. Drawing on original analysis of patent and scientific data, the WIPO patent landscape report on Generative AI provides a snapshot of the patent situation for GenAI.
In the WIPO patent landscape report on Generative AI, discover the latest patent trends for GenAI with a comprehensive and up-to-date understanding of the GenAI patent landscape, alongside insights into its future applications and potential impact. The report explores patents relating to the different modes, models and industrial application areas of GenAI.
1 Generative AI: The main concepts
This chapter provides a summary of the main technical principles around GenAI, including its origins and some historical background. Deep neural networks can usually be adapted to be either discriminative or generative tasks, which has led to the development of various types of GenAI models, which can support different types of input and output data (modes).
2 Global patenting and research in GenAI
Advances in deep learning techniques and increasing computing power have spurred the development of GenAI in recent years. Analysis of patents and scientific publications shows that GenAI is booming across the world and this chapters highlights the top GenAI patent owners, the key inventor locations, and where GenAI patent protection is being sought.
3 Patent trends in GenAI models
Large language models (LLMs) is a term that is now synonymous with Generative AI, generative adversarial networks (GANs), variational autoencoders (VAEs) and diffusion models are all different types of GenAI models. This chapter provides a summary of the main patent trends around different GenAI models.
4 Patent trends in GenAI modes
Generative AI models are very effective for a variety of applications. Mature models support different types of input and output data (modes) and are not limited to text and images, making GenAI potentially relevant to many economic areas. This chapters explores the patent trends for different GenAI modes, from image and video to molecules, genes and proteins.
5 Patent trends in GenAI applications
GenAI is bound to have a significant impact on a wide range of industries as it finds its way into products, services and processes, becoming a technological enabler for content creation and productivity improvement. This chapters identifies 21 application areas in the GenAI patent landscape and explores the different trends in these areas.
The emergence of GenAI will have a significant impact on various industries, providing companies, organizations and individuals with unprecedented capabilities to create, synthesize and manipulate data. However, there are also concerns about the increasing use of GenAI models and tools, ranging from copyright infringement and the potential for misuse, to the risk of job displacement.
Find out more about the methodology used, the approach to searching GenAI patents using traditional patent search techniques alongside new AI search tools, and explore example patents for different GenAI application areas.