AI in the Practice of IP
In the brief time it has been commercially available, AI has had a profound effect on work practices and how firms deliver services. However, professionals should remember the technology is there to assist, rather than take over, as Ralph Cunningham discovers.
One topic has loomed large at this year’s Annual Meeting: Artificial intelligence (AI) and how it is changing and will continue to change the way trademark professionals work.
Brand owners and practitioners alike are excited to explore what benefits the technology can bring to what they do on a daily basis. Speed, efficiency, and consistency of results are all ways in which AI is already removing the burden of mundane, repetitive processes from the shoulders of intellectual property (IP) professionals. How quickly this is happening and how far it has moved into the practice side depends on factors such as where a firm is located, its size, and the nature of its practice.
The technology on which AI is based has existed for some time, of course. Over many years, tasks such as product recommendation, mapping, and fraud detection have all relied on training computers to process large amounts of data and recognize patterns within them.
It is important to note that the term AI covers a wide range of techniques and systems that are used to process data and create new forms of content, and that with every new innovation comes risks as well as opportunities.
Machine learning, where computers learn from data to improve their performance on a specific task; deep learning, which is related to machine learning and involves the use of multi-layered neural networks to analyze various factors of data; and natural language processing, which enables computers to understand and respond to text or voice data in a way that is natural for humans are all forms of AI.
The release of the generative AI software, ChatGPT, in November 2022, brought the technology into the mainstream. Generative AI refers to how computers can be prompted to create a range of content such as text, images, music, video, and software code, based on the information they are trained on.
It became almost instantly successful. In January 2023, only two months after its launch, it had 100 million users when other successful applications such as TikTok and Instagram took much longer to reach the same level of popularity.
According to the latest available data, about 180 million people around the world use ChatGPT. This does not consider the number of people using the competitors that have emerged, including Google Gemini and Claude, and the customized versions that only create content based on proprietary information they are given access to, rather than the whole of the worldwide web.
“These products are alive and available today, but they aren't necessarily all focused on IP per se.”
- Keith Medansky | DLA Piper (USA)
What does this all mean for practitioners?
Already, the advent of AI has produced a number of tools to help them bring speed and efficiency to a range of areas including legal research and e-discovery, drafting, legal analysis and strategy, and enhancing responsiveness. Products such as Casetext’s CoCounsel legal assistant can review and summarize documents, such as agreements, contracts and opinions, search databases, and organize timelines.
“These products are alive and available today, but they aren't necessarily all focused on IP per se. They're focused on things like document review, creating agreements, doing legal research, understanding big blocks of information in a more efficient way,” says Keith Medansky (DLA Piper, USA).
One which is focused on IP is Rocketeer, a tool developed in the London office of Simmons & Simmons. It is trained on a dataset of more than 11,000 EU and 2,000 UK trademark cases, to predict the outcome of a conflict between trademarks in seconds. The firm describes the product as the world’s first AI trademark lawyer.
“Each feature has basically two boxes and a button, so it takes you seconds to use it. The intention is that each time you do so, it's going to save you 20 minutes or half an hour. The end result is that it enhances the way you work, with more data, quicker results and cutting out some of the drudgery and grunt work of your job,” says Darren Meale (Simmons & Simmons, UK), who came up with the idea for Rocketeer.
“Generative AI, the kind popularized by ChatGPT, is a new kind of AI. It changes the game, opening up a whole new set of possibilities but also a whole new set of legal and business issues to deal with in its implementation.”
- Barclay Blair | DLA Piper (USA)
Barclay Blair, Senior Managing Director, AI Product Innovation, at DLA Piper (USA), cautions IP professionals to research and learn about what tools are available in the market and what they do before diving in and spending time and money on products which may not be right for their practice. He says a lot of products are tagged as being AI with no explanation of what this means.
“Generative AI, the kind popularized by ChatGPT, is a new kind of AI. It changes the game, opening up a whole new set of possibilities but also a whole new set of legal and business issues to deal with in its implementation,” he says.
“We are entering a transformative era where AI will be able to do things it couldn't do before and is much more accessible and less expensive for attorneys, without needing a complex technical process to get value out of it. Off-the-shelf, out-of-the-box tools might be great for writing marketing copy but not so great for doing things where precision, accuracy, reliability, stability, and so on, are important, in other words, the work IP attorneys do,” Mr. Blair adds.
Mr. Medansky agrees: “There's a big difference between asking ChatGPT some legal question and getting an answer that could be completely made up versus asking a more controlled platform built to do legal research for lawyers that can give a more reliable result but still requires serious vetting by a lawyer to understand the question asked and any flaws in how it was asked and look critically at the outputs.”
That is not to say that legal professionals are signed up to the uninhibited use of AI in their work all the time. Attorneys from corporations and law firms who responded to a global study conducted by Clarivate in 2023, emphasized liability risk, reliability, and client confidentiality as their greatest concerns about using AI.
“We have got to be able to file, run our processes, meet our clients’ expectations regarding cost, while still being profitable ... That’s where I would say most offices on the African continent are looking to use AI.”
- Herman van Schalwyk | Spoor & Fisher (South Africa)
It is not just the private sector that is using AI technology. The World Intellectual Property Organization’s (WIPO) index of AI initiatives in IP offices around the world shows specific mentions of AI being deployed for tasks relating to trademarks, such as digitization and process automation, image searching, classification, and data analysis. This is only likely to increase.
Source: www.wipo.int
In Africa, private-practice firms are focusing more on how AI can help them with their business processes, rather than on a specific practice like IP.
“We are all looking to streamline our processes. We are looking to drive efficiencies. There’s an element to using AI in our professional practice as well. From an IP perspective in Africa the focus is mainly on the incoming work. We have got to be able to file, run our processes, meet our clients’ expectations regarding cost, while still being profitable. We need to be able to render those services at that price point. That’s where I would say most offices on the African continent are looking to use AI. The term that is broadly used is smart optimization,” says Herman van Schalwyk (Spoor & Fisher, South Africa).
“The legaltech providers are looking to implement off-the-shelf solutions, which hardly fit our purpose, so what we’ve had to do is combine these technologies with our bespoke in-house processes, clearly focusing on particular pain points in our processes,” he adds.
“Document and data manipulation, and associated analytics, are important so we can process volume. That’s really what it’s about. On the professional side, discovery, conflicts, and that side of the work, are getting attention. We haven’t yet developed bespoke solutions that focus on the practice side, because it’s just a small portion of the volume. It’s a bit different for us because we have a lot more incoming work than outgoing work.”
“The reality is we will eventually see the underlying technology used to create very specific tools for specific tasks. That involves getting the right training data and teaching new models for those AI systems.”
- Darren Meale | Simmons & Simmons (UK)
As for the future, will trademark professionals be able to procure AI-driven tools for everything they do?
“The reality is we will eventually see the underlying technology used to create very specific tools for specific tasks. That involves getting the right training data and teaching new models for those AI systems. Then we'll end up with lots of little things we currently do manually, slower, less efficiently, or less cleverly, enhanced by using AI”, says Mr. Meale.
He suspects the larger legal information technology vendors may be less interested in developing these sorts of niche products unless the revenue potential at an enterprise level is clear. This may be a significant factor in the speed of development of AI tools and how soon they become widespread.
“It takes money to develop new tech and create a marketable product, and then money to market it and build a business. Because we're in a niche, with each individual tool doing a little thing, the size of the prize for big tech companies is probably small,” he adds.
Regardless of how quickly the technology progresses, trademark professionals are set to benefit from the ability to do more in less time. These tools will become an important part of practice.
