IP strategy in AI times

By Frank Liu, Shanghai Pacific Legal

The possible impact of artificial intelligence (AI) on the legal industry has come into sharper focus amid the ChatGPT hype. ChatGPT is reportedly capable of collecting, sorting and logically organising data at such a level that many believe it will replace humans in providing legal services. This article shares how AI (including but not limited to ChatGPT) may impact intellectual property legal services and, in particular, strategy.


First, we should be aware of the rapid development of AI and some of its powerful capabilities that extend far beyond those of humans. For example, AI today has proven to be far more efficient than humans in collecting, sorting, classifying and summarising cases, contract templates and other standard-form legal documents that are based on existing databases. Likewise, it is more efficient in analysis, calculation and even some preliminary translations of relevant data. These AI outputs satisfy the “preliminary usability” requirement. AI can even answer some simple legal questions, including those about trademark and patent registration requirements and law in some countries. Such questions will be answered with fast-improving efficiency and accuracy as AI development gathers momentum. But does this mean that AI will replace humans in providing IP legal services and developing IP strategies for corporations?

Frank Liu, Shanghai Pacific Legal
Frank Liu
Shanghai Pacific Legal

Although it is hard to predict AI’s ultimate extent, its instrumental nature will remain unchanged in the foreseeable future. Since the industrial revolution, machines have surpassed all living organisms in power and speed.

Despite some panic in the early days, there have been no changes in the instrumental nature of cranes or locomotives, for instance.

Sure, any powerful enough instrument that is left unchecked may cause disasters. The good news is that humans remain vigilant about AI research and development, as evidenced by an online petition from more than 1,000 tech leaders, including US entrepreneur Elon Musk, calling for a pause in AI training.

We should take a cautious look at AI as a powerful tool with infinite potential, rather than regard it as omnipotent while neglecting its instrumental nature.


How can AI be used to better serve IP strategy-making? We must understand the strong points of both AI and providers of IP strategy-making services. With a combination of their strengths, probably we will become more efficient at developing IP strategy.

For example, before developing an IP strategy for a company, it is necessary to understand the company’s business conditions, its IP profile, relevant legal provisions, cases and development trends. A corporation’s IP data includes but is not limited to trademarks, patents, domain names and business names. Information on its business conditions includes but is not limited to business scope, production and sales and external co-operation.

If they are documented, these basic facts can be summarsied, sorted and preliminarily analysed by AI to save a lot of time. In addition, AI can help greatly improve the efficiency in the sorting and analysis of basic legal provisions, judicial precedents and contracts.

When asked: “What jobs cannot be replaced by ChatGPT?,” the AI chat tool answered: “ChatGPT can perform many tasks that are repetitive, logical and with mass processing of information, but it cannot replace humans where human creativity, judgement and social skills are required.”

While basic facts are important in developing IP strategies, the author believes that a service provider’s experience plays a decisive role in tailoring any IP strategy based on facts and thorough communications with the corporation’s internal staff. In this process, the social skills, judgement and creativity of humans are essential.

It is also the reason why stereotyped thinking cannot deliver tailor-made solutions, as set out in the author’s article “Good communication – the secret sauce in IP protection” (see China Business Law Journal volume 12, issue 8). AI cannot replace humans in negotiation, litigation or enforcement involved in the IP strategy implementation.


Even if AI cannot replace human work in IP legal services or IP strategies, it is certain that we should squarely face the challenges and impacts that AI brings to the legal service industry. In particular, IP services that allow batch processing in a procedure-based and template-based fashion will feel the impact of AI’s strong power in summarisation, sorting, logical analysis and computing.

In addition, AI’s extremely high efficiency and improving accuracy may make it competent for certain work with fixed patterns, such as law retrieval, and case sorting and summarisation, which currently occupy much time for primary legal service providers.

Of course, beginners should still accomplish work themselves to build up their experience and improve their skills, even if such work can be fully taken over by AI. AI may then be used for higher efficiency. Just like the language used for human communication, we need to learn and accumulate a lot from childhood so that we can skilfully use and create in the future.

Going forward, only those who exploit the unique strengths of humans and master AI tools will remain competitive. Those taking stereotypical work roles while neglecting the impact of AI may find themselves at the risk of being replaced by AI.

In conclusion, in the face of tremendous shocks from burgeoning AI, the author, as a provider of IP legal services and maker of corporate IP strategies, believes that we should not underestimate the advantages of AI unattainable by humans, nor stick to the old ways, nor fear the power of AI. Instead, we should prudently harness AI as a powerful instrument. The new era will ultimately belong to lifelong learners who keep learning and master powerful AI tools.

Frank Liu is a partner at Shanghai Pacific Legal

Room 2709, 27/F, Plaza 66 II
1266 Nanjing Road West
Shanghai 200040, China
Tel: +86 21 6086 0199
Fax: +86 21 6086 0111


Back to INTA2023 page for more IP articles