questel_logo
closeClose
More
EN
questel-menu
What does patent data reveal about the race to patent AI, image of 3D white blocks
Landscape / Published, November 20, 2024

A Revolution in Progress: What Patent Data Reveals about AI Innovation in IP

overview

Artificial intelligence (AI) is radically transforming all business sectors, bringing innovative solutions and opening new horizons—and the IP field is no exception to this trend, as evidenced by our analysis of AI-related patent filings. Our research reveals that the number of patents filed since 2013 is increasing exponentially, with the patent data showing that AI is being deployed through various applications in the IP field. 

Earlier this year, the Questel IP Consulting team undertook a patent data analysis of AI innovation in the IP sector for our Industry Outlook Research 'Beyond the Hype: How Technology is Transforming IP'. Here we summarize some of the key findings from that analysis and what the race to patent AI in the IP field means for IP professionals.

On the Rise: Patent Families for AI in IP

Our patent data analysis reveals a compound annual growth rate (CAGR) for AI-based innovations for IP of 30% between 2013 and 2023: 

Chart showing key patent AI trends from our patent data analysis of filings for AI-based innovation in IP
©Questel 2024

Who Is In the Race to Patent AI?

 

This boom in filing activity is the result of a multitude of players from different backgrounds seeking to patent AI technologies:  

  • Specialist IP providers (established players) are responsible for the majority (26%) of AI in IP filings but at a lower level compared to new entrants (15%).  
  • IT companies (24%) are developing specific systems, such as image recognition, adapted to IP.  
  • The patents filed by consulting firms (12%) focus on the financial valuation of IP and tools to improve state-of-the-art search productivity.
  • Large corporations (10%) and law firms (6%) are protecting methods and solutions—probably more for defensive purposes than product development. 
Pie chart showing who is leading the race to patent AI, from our patent data analysis of AI-based innovation

The AI Technologies Covered by Our Patent Data Analysis

As part of our research, Questel's IP Consulting team took a closer look at nine promising technologies:
1. Interactions with patent and trademark offices (PTOs):
Generative AI applications that assist in patent and trademark prosecution, e.g., submitting documents and responding to office actions.
2. Landscape, patentability, clearance, and invalidity search:
AI tools designed to automate patent, trademark, and design searches, including for patent data analysis.
3. Report drafting:
AI-driven solutions that compile and draft comprehensive reports, such as patent landscape analysis, to provide insights into the state of the art in a particular technology domain.
4. Patent drafting:
Generative AI systems that assist in drafting patent applications, including claims, detailed descriptions, and embodiments of the invention.
5. Identifying licensing opportunities:
AI technologies that analyze patent portfolios and market data to identify potential licensing opportunities.
6. Patent database improvement and categorization:
AI algorithms that enhance current patent search capabilities, allowing for more efficient identification and categorization of patents.
Discover Questel's AI-elevated tools for IP
7. Advice chatbots:
Interactive AI-powered chatbots that provide guidance and answers to common questions in the patent and trademark fields.
8. Image search (patent, trademarks & designs):
AI-enhanced image search technologies tailored for identifying and analyzing visual elements in trademarks, patents, and designs.
9. Patent translation:
AI systems specialized in translating patent documents across multiple languages, maintaining the precise and technical nature of the language used in patents.

AI and Maturity: The Hype Cycle

The Gartner Hype Cycle provides a helpful way to visualize the process a new technology goes through as it enters the mainstream. It enables businesses to understand where a new technology is in its lifecycle and to make informed decisions about when or if to invest in it.  

  • Where Does AI for IP Sit On this Hype Cycle?  

Gartner hype cycle showing patent AI trends from our patent data analysis 

icon
Stage 1 ’Technology Trigger’—A new technology sparks interest and initial media coverage.
chevron
  • (5) Identifying licensing opportunities: A niche application of AI with potentially high value focused on identifying the best candidates in an IP portfolio and searching for licensees and/or evidence-of-use on the market. However, likely in a very early stage.
  • (1) Patent and trademark prosecution or other PTO interaction: This application is emerging, with potential breakthroughs being explored. The use of generative AI for navigating the complexities of patent and trademark prosecution, including interaction with multiple PTOs is still in the early stages. Interest is growing as more proof-of-concept applications are developed.  
  • (8') Image search (Patents): A promising way to complement a more traditional search based on keywords, classifications, citations, and semantic analysis... The field remains at an early stage because of the complexity of interpreting images according to the patent description.  
icon
Stage 2: The Peak of Inflated Expectations—As stories about the potential of the technology spread, excitement builds.
chevron
  • (7) Advice chatbots for IP: While chatbots are being used for customer service, their application in providing legal or search advice is still in development or just on the market for simple tasks.  
  • (2) Landscape, patentability, clearance, and invalidity search: A promising area with technology that is already used to help search and retrieve relevant documents. The ability to handle complex legal issues and to search effectively using simple words is still to be validated. 
icon
Stage 3: The Trough of Disillusionment—When the initial promises aren't met immediately, enthusiasm wanes and the technology faces criticism.
chevron
  • (3) Report drafting: Report writing associated with a patent portfolio or landscape faces difficulties related to legal nuances and complex patent analysis. Likely, it will not entirely replace human expertise.  
  • (4) Patent drafting: Initial enthusiasm around the potential to use generative AI for patent drafting has faded due to difficulties managing complex legal and sector-specific concepts and nuances. Now, the emphasis is on using AI as a tool to enhance and speed up human patent drafting.  
  • (6) Patent database improvement and categorization: Many platforms now offer sophisticated, AI-driven search and categorization tools that enhance productivity. However, effectiveness and accuracy are still being challenged depending on the field and precision needed.  
icon
Stage 4: The Slope of Enlightenment—As the technology matures and practical applications are developed, interest starts to climb again.
chevron
  • (8) Image search (Trademarks & Designs): AI-powered image search is now a well-established and efficient technology and is already available in a few trademarks and design search tools.  
  • (9) Patent Translation: Machine translation with AI assistance is used frequently. The quality obtained with a 100% machine translation is acceptable for patent databases, while human expertise and review are still needed to translate patent applications for filing.  

The Race to Patent AI

: Will Emerging Technology Replace IP Expertise?

Over the last few years, the IP sector has begun exploring AI’s potential to provide new tools, features, and functionalities across many different fields. As shown by the patent data and patenting dynamics in this field, research into new features and improvement of existing ones remains incredibly dynamic.  

Now the technology has been unleashed, this enthusiasm will not stop; however, we expect greater investigation, knowledge, and application to build a general understanding that AI-enabled technologies for IP are progressive rather than "magical" tools. In other words, we expect AI to improve and accelerate human work but not supplant human expertise. However, while it is likely that AI will not replace IP practitioners, those who do not use AI will most likely be replaced by those who do. 

Discover more insights on the impact of AI technologies on the IP sector by downloading our Industry Outlook Research 'Beyond the Hype: How Technology is Transforming IP' or contact our IP Consulting team for additional insight and support.

About the Authors
null Cyril Mavré
Cyril Mavré
is Patent Consulting & Prior Art Director at Questel.
null Xavier Dieumegard
Xavier Dieumegard
is a Practice Leader in the IP Consulting team at Questel.
Talk to our experts
Discover our plans for AI and IP and join us on this journey by contacting our team today.
Contact us
Discover our patent data and patent AI services, image of building