The Real Story of How Researchers Can Use Intellectual Property as a Weapon - In the AI Era, I Asked 20 Veterans About the Skills Needed 'Before Qualifications'

Will AI steal our intellectual property jobs?

Entering 2026, generative AI has made a sudden, massive entry into the field of intellectual property. AI handles prior art patent searches and drafts for specifications in the blink of an eye. So, will the opportunities for us—researchers and those aiming for IP careers—diminish?

If you had asked me this six months ago, I wouldn't have been able to answer well. But now, I can say it clearly.

Last year, I dove into a JIPA (Japan Intellectual Property Association) social gathering, the only one among 200 people, and went around asking 20 veterans the same thing: 'If a researcher is moving into intellectual property, what is needed first?'

That answer carries even more weight now in the AI era. Today, I'm going to talk about that.

What you will learn in this article
- The skills truly required in the IP field before a patent attorney qualification
- My current first-hand experience trying patent searches with generative AI
- Raw, practical stories that I couldn't learn from studying for the Intellectual Property Management Skills Test Level 2

The 'one thing' 20 people agreed on before a patent attorney qualification

(What you will learn in this chapter: The story of how it's not qualifications that are truly tested in the IP field)

'You can't do IP without a patent attorney qualification, right?'

This is the question I always get when I say I'm interested in IP. I used to believe that, too.

But after going around and asking 20 veterans at the JIPA social gathering, that assumption vanished.

Everyone said the same thing, like they were reading from the same script.

'Skills can be learned later. What you need first is communication skills and initiative.'

Honestly, I was let down.
(Not a single person mentioned the name of a qualification.)

There are three main ways to work in IP.Patent firms,Corporate IP departments, andIP consulting. Among these, a 'patent attorney qualification' is a prerequisite when acting as an agent to handle procedures.

However, if you are going to use your research experience as a weapon to work in a corporate IP department, a qualification is not an entry requirement. What is asked for first is the ability to talk to inventors and a proactive attitude.

Why the value of those words has 'risen' in an era where AI reads patents

(What you will learn in this chapter: The more generative AI spreads, the clearer the skills that remain for humans become)

Six months have passed since then. Generative AI has become a standard part of the IP field. Searching, writing, summarizing. AI is clearing things away one after another.

So, have the words of those 20 people become outdated?

It was the opposite.

As AI takes over 'researching and writing,' what remains for humans isthe ability to draw information out of inventorsandthe ability to translate that content into language that management can understandbecame more important.

A center director at a precision equipment manufacturer once said this:
“Digging out inventions is a human job.”

In this AI era, those words have hit the mark.

I have personally experienced consulting with a patent attorney on 'how to file an application to best protect this technology,' and having them ask, 'Do you have any more information like this?' to draw out details I hadn't even realized myself. That process of digging up information through repeated dialogue is something AI cannot replace.

The more convenient tools become, the more human-to-human dialogue matters in the end. The 'communication skills and initiative' that 20 people mentioned six months ago have not expired; in fact, their value has increased.communication skills and initiativehave not expired; in fact, their value has increased.

Tips I learned from trying patent searches with generative AI, and my challenge with the JPO API

(What you will learn in this chapter: Pitfalls when reading patents with generative AI, and the tricks I am currently testing)

From here on, this is my 'right now' first-hand experience.

After trying patent searches with generative AI, I felt the biggest strength was 'time reduction.' Both reading and organizing are clearly faster.

However, I hit a wall immediately. General-purpose generative AI cannot directly pick up patent information from J-PlatPat (the Japan Patent Office's free search database). This is because J-PlatPat is designed to be operated only from a browser screen. I can barely pick up information from Google Patents, but that doesn't cover everything either.The information you can retrieve changes drastically depending on how you conduct the search.

That is when I learned about thePatent Information Retrieval APIthat the Japan Patent Office is providing on a trial basis. It is a system that allows you to mechanically retrieve application progress information and the like in data format using the application number as a key, and you can register to use it even for personal research. I have applied for this and am currently trying various things to see if I can make my research more efficient.

However, do not have excessive expectations. This API is still in the trial stage, and the information you can get is mainly progress information. You can only find information for one application number per access, and there is a limit to the number of accesses. Since the specifications and registration windows can change, if you want to use it, please be sure to check the latest status on the Japan Patent Office website.

In other words, it is not the case that you can pick up everything if you use AI.What to leave to AI and what to verify with your own eyes. Drawing that line will be the key to patent searching from now on.

(I would like to summarize these practices in more detail in another article once they take a more concrete shape.)

'Date management' that I didn't learn in the Intellectual Property Management Skills Test Level 2, and the story of 'selective' intellectual property

(What you will learn in this chapter: The gap between textbooks and the field, and the ideal state of an IP department)

I studied for the Intellectual Property Management Skills Test Level 2 and acquired a solid foundation in the law. But the field was, as expected, a different story.

The most effective thing was the sense ofdate management.

About a month before a certain exhibition, I learned this for the first time.

If you disclose an invention at an exhibition before filing for it, as a general rule, that invention loses its novelty and you will be unable to obtain a patent.

There is a system called the 'Exception to Loss of Novelty' (Article 30 of the Patent Act) to provide relief. However, the procedures are extremely strict—for example, you must state your intent in the application at the time of filing and submit a certificate within 30 days of the filing date.

If even one step is missed, you lose the rights permanently. Moreover, in many countries overseas, this exception is not recognized at all.

At that time, the patent attorney used their quick wit to organize the application documents so that they would meet the deadline. I felt a chill run down my spine.

Since then, I have made sure to keep track of which tasks need to be completed by when, so as not to leave the entire schedule up to the attorney. In practice, because you file through a patent attorney, the deadlines are set with more buffer than what is written in textbooks. Therefore, when I receive requests for filings or documents, I return them as quickly as possible. That has become a habit before I knew it.

This sense of 'taking initiative' was directly connected to the proactivity that the 20 veterans spoke of.

And I began to think this as well.

Even within intellectual property, there are departments that just handle assigned tasks, and departments that can use the information they gather to influence management strategy.

The latter is what I want to choose. Each member of the department interacts with the outside world on their own and gathers information themselves. It is in such an IP department that the technical eye cultivated in research becomes a true weapon.

It's not a binary choice between continuing as a researcher or quitting

(What you will learn in this chapter: How intellectual property is a weapon that works for both those who quit research and those who continue)

If you have read this far and interpreted this as a story saying, 'In short, quit research and go into intellectual property,' that is a little different.

The ability to read patents itself is a weapon for those who continue in research as well as those who move on to a different path.

To protect your own technology and to avoid inadvertently infringing on the rights of others, engineers should know about intellectual property. That was the starting point for why I wanted to learn about IP in the first place.

If you are going to live as an engineer, you should first step onto the playing field of 'intellectual property' that handles technology with your own two feet.

That night at JIPA, one veteran said this to me:

'If you come with that attitude, any IP department will be happy to have you.'

I still remember those words.

Finally, just one thing. Next week, I will be taking a break from updating the blog to devote all my energy to the Intellectual Property Management Skills Test Grade 2. I am going to give everything I have built up so far.

Not qualifications, not AI, but the ability to take initiative. Since that night half a year ago, the assumptions I held have certainly vanished.

So, I will answer this question one more time and then put down my pen.

"Will AI steal the jobs of intellectual property professionals?"

——I believe it will not.

💬 Please let me know in the comments if there was a moment while working as a researcher when you felt "conscious of intellectual property,"
or if there is something you wish you had known sooner.

Please let me know in the comments.

▶ If this article was even a little bit helpful, it would encourage me if you could support me with a "Like" and a "Follow."

📚 Recommended reading