This study forms part of a wider research piece ‘Can the UK win the global R&D race?‘. Material is available to re-publish, with credit (including link) to Source Advisors. This report was published on 29 July 2024. Please get in touch for additional information.
In healthcare, AI-powered diagnostics and predictive analytics improve patient outcomes by enabling early detection of diseases and personalised treatment plans. In finance, AI-driven algorithms provide more accurate risk assessments and fraud detection, enhancing security and decision-making. Additionally, AI facilitates advancements in transportation through autonomous vehicles, reducing accidents and optimising traffic flow.
AI is pushing the boundaries of what machines can achieve. Advances in machine learning, natural language processing, and computer vision are enabling systems to learn, understand, and interact with the world in ways that were previously unimaginable. These breakthroughs are not only driving efficiency and productivity but also opening new possibilities for addressing complex global challenges. As AI continues to evolve, its potential to reshape our future becomes increasingly evident, making it a critical area of focus for researchers, businesses, and policymakers alike.
Patent filings is a good indicator of technological growth. Usually, an inventor will file a patent application to protect and commercialise their technology. Granted patents offer the owner rights to stop others from exploiting the patented technology (use, sell, make, distribute) for up to 20 years. To qualify for patent protection, the technology must be novel and considered inventive according to local patent law. Usually, only one patent is granted per invention and therefore the number of patent applications is an indicator of the number of proposed inventions for a given technology area.
Figure 1 shows the level of growth in patent filings for the AI industry. These figures are the most dramatic yet with close to a 2000% increase in filings for the period between 2010 and 2022 with growth continuing to skyrocket to over 300k filings in 2026 alone. Now, this is a high growth industry with even greater growth possibilities, and we can see that in our everyday lives, automation/artificial intelligence is playing a role in all aspects, from providing better business insights (such as this article) to novel drug discovery. The potential of this industry is much higher than the current technological baseline and therefore commercial opportunities are rife and wide ranging.
Figure 1. Volume of patents filed within the AI industry from 2010-2024 and projected growth figures for 2024-2026.
Geographic breakdown of AI patents
Given that patent applications are territorial, we can trace back the origins of a technology by the first (priority) filing within a patent family. For reference, a patent family usually relates to a single invention and the family members are represented by the number of countries (foreign filings) that invention has. Usually, the first filing is made in the country of origin and the chart below (Fig.2) shows the percentage (%) share of territories where first filings are made for inventions within the automotive industry. This chart illustrates all filings made from 2010 onwards that are granted and active or currently patent pending, therefore excluding any expired or lapsed patents or patent applications.
China is very much dominating the percentage share of filings with close to half of all global filings related to AI. Therefore, even with high filing volumes related to automotive and telecom technologies, AI is far and above their number one strategic priority. The US follows in second with a ~30% share of all filings giving China and the US a greater than 75% share in the number of global patent filings (see Fig.2).
The UK by comparison has relatively fewer filings with a ~1% share. It must be said that the US and Chinese patent system is more AI/software friendly, particularly when it comes to technologies related to improving business methods. By comparison, such technologies are not patentable in the UK and in Europe by law.
Figure 2. Geographic breakdown by percentage share of patent filings for the AI industry between 2010-2024.
All patent applications are tagged with technology indicators to categorise filings by its constituent technologies. I say technologies, rather than technology, because it is very common for a patent application to have more than one technology indicator. Usually new technologies are born from a mix of existing technologies or at least are classified to a certain discipline of science and technology.
Given that patents are classified in this way, it is possible to search and filter for patents that contain technologies related to the AI industry.
Figure 3 gives a useful breakdown of the technologies developed within the last 14 years. Biological models, related to advances in neural networks such as recurrent neural networks (RNNs) and convolutional neural networks (CNNs), is where most filings are concentrated. Followed closely by technologies to improve character and pattern recognition, in addition to image analysis and computer aided design (CAD).
Figure 3. Technology breakdown (AI) by percentage share of patent filings since 2010.
So, who is driving growth in these technology areas? Figure 4 illustrates just that by overlaying the technologies by country of origin. We have removed China from the graphic below as the data skewed heavily in favour of China based applications.
The US dominates most of the top 10 technologies in the AI space, particularly specialising in neural networks, data processing and machine learning. Korea and India follow suit with influence over character and pattern recognition technologies and image analysis.
Figure 4. Technology breakdown by origin country for the AI industry.
The UK picture is slightly different in the types of technologies filed most when compared to the US. Like the US, neural networks come out on relatively high, however, the UK appears to have specialisation in securing communication specifically with filings driven by NChain. Image analysis and image enhancement are also specialist areas, driven by filings from Sony Interactive Entertainment and Imagination Technologies (see Fig 5).
Figure 5. Technology breakdown by volume of filings originating from the UK for the AI industry.
About Akshay Thaman | IP Consultant & Policy Lead
Akshay works closely with Source Advisors’ clients to uncover IP that may be hidden within their businesses. As a Member of the British Patent Information Professionals group (BPIP) he brings academic rigour and commercial experience to his role as IP Consultant & Policy Lead.
View profile
Data collection for this report was conducted on 21 June 2024 for patents filed from 1 January 2010 onwards. This report analyses filings that are either granted and active or currently patent pending, therefore excluding any expired or lapsed patents or patent applications. Read our full report ‘A focus on key technologies – how does the UK compare to other global leaders?‘.
