I started working as a part time researcher in a company during my master degree as part of my graduate studies. I published a paper 2.5-3 years ago based on the data and my work at school. Last month, I noticed that a competing company patented the idea that was presented in my paper in 2019 with minor twisting.

Can anyone patent my idea from my published research paper? I integrated a new idea an AI system 3 years ago. When I discussed this with the company that I work with, they said that:

Firstly they claim that they are surprised that it is patentable. They said even we did the work, they might get the patent on that particular aspect because we did not publish anything or we did not apply so whoever applies first gets the patent. We do not know if it is stolen or came up with it independently. We will talk about it in the next little while. We need to spend some time on, we do not see any urgency for taking action. We can not supply information to the patent examiner that will invalidate their patent. We need to be careful of how we do that because we can go through a bunch of work and provide them with some information. If the patent examiner still gives them a patent, we have done some work to makes them strengthen their position. We want to be careful of how we approach it.

I am not happy about this and got very disappointed on how they did not protect my work and my research. Do you think does it worth to do further research in this company? Or should I focus my academic aspects of work outside of this company? For example asking for help from my university? I just graduated 5 months ago and working at the same place. My supervisor also claims that he did not know it would be a patent. He says that he would like to follow company's approach.

But my view is that the company is not taking the right action as they only care about the sale and profits rather than research. Please advise! What is the best action for me?

I am not looking for a legal advice, what I would like to know" if I should keep doing research here or I move on?"

  • 3
    I don't work in a field where we generate patents, but I'm confused how someone patented something that you already published on? I'm afraid I'm having trouble understanding the quote, is it translated? Was the patent in a different country than you're in? Commented Feb 14, 2022 at 21:27
  • 2
    In the US you probably have no recourse. But you'd need to talk to a patent attorney to know. You need to file a patent early to get any protection. Since it is a legal question, I'll vote to close.
    – Buffy
    Commented Feb 14, 2022 at 21:32
  • 1
    This seems like a legal question between your employer and a competitor of theirs. If you or they want to challenge a patent, you'll need legal advice on how to proceed. I don't think Academia.SE can help you with that; there doesn't seem to be anything you need to do from an academic perspective.
    – Bryan Krause
    Commented Feb 14, 2022 at 21:45
  • 1
    Why is not patenting something a cause to move on? Patents cost money (to prepare, file, and maintain) and few actually cover their costs. Should the other company sue, there is your previous art to invalidate perhaps some portions of it.
    – Jon Custer
    Commented Feb 14, 2022 at 22:37
  • 4
    Note that patent generally doesn't apply to "ideas". It applies to "useful devices". Idea needs to be realized in a "device" to be patentable. More generally "ideas" have no ownership rights. Different sorts of realization result in different sorts of IP rights.
    – Buffy
    Commented Feb 14, 2022 at 22:48

6 Answers 6


Other answers correctly point out what you, yourself, should do to invalidate the patent (nothing).

Your question also asks (paraphrased) "should I quit over the company’s (lack of) action?". Only you can answer that; I will try to help your decision-making by explaining why, at this point in time, the company’s decision to do nothing seems entirely reasonable.

The following is written by someone whose entire legal education is a boring 20h training in patent law, and who does not know the specifics of your case. If you rely on this for legal action, you deserve whatever happens to you.

The procedure for patents in most Western jurisdictions (US/EU) is as follows: once you file an application, the patent office will examine it, and make their own search for prior art. After some delay (18 months in Europe), the patent application as well as the prior art search report are published (visible to everybody). After some more delay, during which competitors may object, the patent is granted (maybe partially, if some claims were invalidated by prior art etc.). For more details, Wikipedia has an article about patent prosecution.

(Notice also that stuff you published yourself can be part of prior art and invalidate a patent that is filed later. If you wanted a patent for yourself, the time to act was before you published an article, not now.)

A competitor may attempt to invalidate a patent after it has been granted. It is not uncommon to gather prior art evidence as soon as a competitor’s patent is granted, so that the legal department has a better idea of the risk of litigation, but wait until an infringement procedure is launched to actually do anything with it. Some legal proceedings require you to put up objections as soon as you are aware you can make them, but patent application is not one of them.

From what you say, I assume the patent was just published, and is not granted yet; that you published your paper before the filing date of the patent application (18 months back in time); and that you want your company to object to the patent while at the patent office stage.

Objections at the patent office are (much) cheaper than litigation at court. However, no litigation whatsoever is yet cheaper than objections at the patent office. If your company does try to sell stuff that falls within the patent claims, and the competitor tries to enforce the patent (via a lawsuit), the opening move from your company’s patent attorney will likely be to write a letter saying "here’s the prior art, are you sure you wanna do this?" (paraphrased). If your publication is crystal-clear prior art, such a letter may be enough. That letter may still be somewhat expensive (because the lawyer will have to interview you and check records before writing it), but it will still be orders of magnitude cheaper than actual litigation.


There are a few points to consider here.

  1. Prior art does not necessarily invalidate a patent. It may invalidate some claims, but other elements of the patent may still stand.
  2. Your thesis may or may not constitute prior art. This is particularly true if your thesis was published at the same time as the patent.
  3. Litigating a specific patent doesn't necessarily do anyone any benefit, and will possibly involve non-trivial costs.
  4. Your point about "care about sale and profits rather than research" somewhat contradicts your entire post; if they cared only about research, then they almost definitely wouldn't care about the patent.

All that said, the answer to the question, "what should I do?" is likely "nothing". Unless you specifically want to pursue an idea that revolves around having patent rights to this idea—which is really something that only a patent attorney can address—you should probably just move on.


There are estimates that in the software realm, five out of six patents are not actually enforceable because there is prior art that mean that the patent should not have been granted in the first place. (I will assume that your artificial intelligence work falls in the "software realm" of patents.) Indeed, it is exceedingly rare that companies have successfully used software patents to force a competitor to stop making a product, or to force them to pay royalties for a patent the suing company owns. The usual strategy is simply to intimidate a competitor by threatening to sue over a whole portfolio that may contain thousands or tens of thousands of software patents; that will cost competitors years of money and uncertainty, but the validity of the patents will never actually be put to the test. It is commonly believed that if companies ever tried to really use their patents, most of these patents would probably be invalidated because the sued party could show that there is ample prior art.

For you, this really just means that you could spend a lot of money trying to challenge the granting of the patent, or try to get it invalidated retroactively. Or you could say "who cares; they can't do anything with the patent anyway". The latter option costs you nothing, and it doesn't even cost you any opportunities: You can continue to use your work as if there was no patent, because the only entity that could sue you would be the company owning the patent, and they will retract any threat of a law suite once you show them that you produced prior art that would invalidate their patent if they tried to enforce it. That is also the reason why the company you worked for doesn't care: They, too, know that that patent is easily invalidated, but it's not worth anyone's effort to do so.

In other words: Move on with your life. It's not worth anyone's effort to deal with it.

(I will add that you can not patent the work yourself either because you published on it. You cannot patent something you have already published. You would have had to patent it first, but that train has left the station.)


I'm mostly responding to your comment that you are interested in the future, not the past.

First, I think the company's response was pretty accurate if they didn't think it was patentable and you didn't convince them otherwise. The advice about the difficulty of a challenge and the likely outcome seems accurate.

I don't know what sort of "patent sharing" agreement you have with the company. If it is ad-hoc, then you have a problem. But many companies will bear the expense of a patent and the possible expense of a challenge, contracting to pass some of the financial benefits to the people who did the work. If you did the work on "company time" it is unlikely that you can patent things on your own. So, in the end, you need to convince management that the cost/risk is worth bearing.

But, make sure you have a written agreement in advance and that the company has, in the past, behaved properly in such situations.

If you move on, and continue to do the same kind of research elsewhere, then you will have exactly the same issues. You and the company need to agree on the risk-value equation since the financial risk is mostly theirs.

And, patenting as an independent researcher can be very expensive, not just for the initial filing, but for the legal cost involved if you get challenged.

But, in the end, I think, everything flowed from their initial reluctance about the patentability of the "device" and the financial risks and rewards. If you had solved that, this wouldn't have been as likely to happen, unless the other company was doing parallel research (or industrial espionage).


What is the best action for me?

I know this can be frustrating, but most likely, nothing can be done realistically, except from being extra careful for the future inventions. The intellectual property one creates as a part-time researcher in a company is typically owned by the company. Therefore, it's up to the company to decide what to do with it. If someone patents your paper, then it can be invalidated (mirror) if one can prove the following, but your company seems uninterested to spend time and money it:

Another approach to invalidating a patent is to prove that the invention was in public use or on sale in the U.S. more than a year prior to the date of the application. If one can prove that the invention was known or used by others in the U.S. or patented or described in a publication in any country before the invention thereof by the applicant, the patent is invalid. In order to obtain a patent, the invention must be useful, novel and unobvious.

You could try yourself but it'll likely be time-consuming. The quote above is for the US, each country has their own patent system/rules.

Note that publishing patentable work is referred to as a defensive publication:

A defensive publication, or defensive disclosure, is an intellectual property strategy used to prevent another party from obtaining a patent on a product, apparatus or method for instance.

  • 1
    You can’t patent a “paper”, actually. Devices are patentable.
    – Buffy
    Commented Feb 14, 2022 at 23:10
  • @Buffy thanks, "patent a paper" was a shortcut to say patent the content (eg, a novel algorithm) of the paper, assuming it's patentable. Commented Feb 14, 2022 at 23:16
  • Further, no one can "patent the content of your paper". They can patent their own work, and this work could overlap with what's in your paper, but they aren't patenting your paper or your content, they're patenting theirs.
    – Bryan Krause
    Commented Feb 14, 2022 at 23:39
  • @BryanKrause Thanks, yes that's what I meant. OP's pb is that the content is very similar. Commented Feb 14, 2022 at 23:42

Sorry this happened. But, you’re the one and only who really knows the project. I would start looking for how to modify it in a way that it would be even better and patent it in a hurry, before the others do it. The improvements must be significant to be valid. You have now the advantage of being able to see the results of the first device that was produced and, you can look for it’s flaws.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .