China is spearheading the field of generative artificial intelligence (GenAI) patents, according to a new report by the UN World Intellectual Property Organization (WIPO).
The report reveals that from 2014 to 2023, Chinese inventors filed over 38,000 GenAI patents. This figure is six times higher than the patents filed by their counterparts in the United States, which holds the second position.
RAPID GROWTH IN PATENT FILINGS
Since the introduction of the deep neural network architecture in 2017, synonymous with GenAI, the number of patents in this field has surged by over 800% through 2023, according to the WIPO report.
The dramatic rise in patenting activity highlights the recent technological advancements and the immense potential within the GenAI sector.
Daren Tang emphasized that by analyzing patent trends and data, WIPO aims to provide a clearer understanding of the development and future directions of this rapidly evolving technology.
WHAT IS GENERATIVE AI?
Generative AI, often abbreviated as GenAI, is a transformative technology that enables the creation of content, including text, images, music, and software code. This technology powers a wide range of industrial and consumer products, such as chatbots like ChatGPT, Google Gemini, and Baidu’s ERNIE.
“GenAI has emerged as a game-changing technology with the potential to transform the way we work, live, and play,” said Daren Tang, WIPO Director General.
SHAPING THE FUTURE
“This can help policymakers shape the development of GenAI for our common benefit and ensure that we continue to put the human being at the center of our innovation and creative ecosystems,” Tang stated.
TOP ORGANIZATIONS WITH THE MOST PATENTS IN GENAI
1. Tencent
2. Ping An Insurance Group
3. Baidu
4. Chinese Academy of Sciences
5. IBM
WHERE ARE THE MOST GENAI TECHNOLOGIES INVENTED?
1. China
2. United States
3. Republic of Korea
4. Japan
5. India
6. United Kingdom
7. Germany
WHICH GENAI MODEL HAS THE MOST PATENTS?
1. generative adversarial networks (GANs)
2. variational autoencoders (VAEs)
3. decoder-based large language models (LLMs)
