Is it worth patenting an algorithm if I don't have the money to defend against infringements?

29

10

I have some algorithms that I am considering patenting, but I wonder whether I should just trust that rivals won't simply read the patent papers and copy my ideas.

While I hope that my inventions will soon improve my financial circumstances, I currently do not have the time, money or skills to determine whether another commercial software application is stealing my ideas. Their software would be closed-source, so the only ways to verify that they are infringing my intellectual property would be to either to seek a whistle-blower within that organisation or to reverse-engineer their application. And then I still have to come up with the money to sue!

Do the pros of having a patent outweigh the cons?

DonkeyBoy

Posted 2017-05-24T08:15:56.290

Reputation: 413

2Btw. congrats on getting an ask patents question into the hnq list! – DonQuiKong – 2017-05-24T15:48:29.270

3You might consider keeping the algorithms as a Trade Secret. – Eric S – 2017-05-24T17:07:00.153

2@EricShain I was thinking about doing this, but I've recently been advised that large corporations have the resources to reverse engineer software, to find out how it works. So a trade secret might not stay a secret for long anyway. It's a tough decision! – DonkeyBoy – 2017-05-25T02:51:35.010

1

Have you spoken with a patent lawyer who specializes in your field? It's not uncommon for a first consultation to be free. One option is to file a provisional patent, which can help establish precedence yet allow you time for further work. https://www.uspto.gov/custom-page/provisional I would have other questions about your plans for revenue generation and/or plans to find investors, both of which are relevant to the discussion.

– Rethunk – 2017-05-25T16:20:34.137

@Rethunk I did actually have an initial consultation but part-way through, I lost faith in the lawyer because he kept insisting that I pay a full patent fee for the provisional and vice versa. He tried to make it sound like it would be to my advantage, but when I pressed him on the point, he was forced to admit that his motivation was purely to benefit himself only. I just don't trust any particular lawyer well enough to be sure that this won't happen again. – DonkeyBoy – 2017-05-26T01:40:02.140

@DonkeyBoy. Sorry you had such a bad experience. You could network a bit to find a better lawyer--and they're out there--but I understand your reluctance. You could write a provisional yourself, submit the fee (about $150) and would have a year to decide whether to pursue it, during which time you might secure funding for a full application. If you can secure funding and/or make some significant initial sales, then a patent would be worth it. Otherwise I'd suggest using code obfuscation, adding unique fingerprints to your code, and focusing on revenue at first. – Rethunk – 2017-05-26T16:03:19.150

Answers

22

This is a great question, with no one right answer.

The strength of a patent comes from the ability to enforce it. If the patent owner cannot enforce it for whatever reason (such as a lack of funding), the patent is effectively toothless.

But your competitors will likely not know that. They may therefore be dissuaded from infringing simply by the existence of a patent and the assumption that you would sue. Such a chilling effect on competition may be valuable to you.

In addition, if there is infringement, you may be able to get a licensing agreement without the need for litigation. This can be a very lucrative business model, and so may justify the cost of patenting.

And if litigation seems likely, there is a nascent field of litigation funding. While your funders will take quite a great deal of the damages, you may at least get something out of the end of it.

Finally, even if you can't sue, someone else could. You might therefore get quite some value from selling it off at some point. There are a number of tech businesses where most of their value comes from their IP portfolio.

On the whole therefore it depends how you see yourself proceeding. If you're unsure, it may pay to err on the side of patenting: once you disclose or use your invention publicly, you're generally barred from patenting it later (but for a short grace period in some countries), whereas it's easy to abandon a patent (or application) later down the track if you wish.

Maca

Posted 2017-05-24T08:15:56.290

Reputation: 6 985

4To emphasize the selling point, patent trolls often buy good patents - while I personally don't agree with their business assumptions, when it comes down to either not being able enforce a patent or sell it to a troll, I'd advise the latter. – DonQuiKong – 2017-05-24T09:43:00.853

5There are also attorneys (at least in the US) that take on cases on a contingency basis. That is, they only get paid if they win. Of course if the win they get paid a significant percentage of the award. – Eric S – 2017-05-24T11:13:17.173

2Also, the existence of a patent would generally prevent an honest person from infringing it (if they knew of the patent). And you might have more funding to enforce it by the time someone infringes it. – John B. Lambe – 2017-05-24T12:24:30.260

2"But your competitors will likely not know that. They may therefore be dissuaded from infringing simply by the existence of a patent and the assumption that you would sue. Such a chilling effect on competition may be valuable to you." A patent bear's the applicant's name, and generally it would be easy to find out quite a few details about the applicant's business (if any), funding, work experience, prominence in the field, and so on. This assumes the application survives examination, and also that a competitor sees any value/threat in the patent. – Rethunk – 2017-05-25T16:29:26.210

2Also, this strikes me as unclear and possibly misleading: "you only get one chance to patent something before it's forever unpatentable" The timing of a provisional application and/or full application and the time when a product is offered for sale allow more than "one chance" to patent an invention. – Rethunk – 2017-05-25T16:37:17.083

1@Rethunk A fair criticism, which I've adjusted my answer to address. – Maca – 2017-05-29T10:08:52.230

Isn't it true that, once you put a company on notice of infringement, you start racking up damages? Can someone confirm this? So even if you don't have the money to sue today, this does not mean that you sacrifice your ability to seek those damages later down the road. – jdpatent – 2017-05-30T23:16:36.303

@jdpatent It's a little more nuanced than that. Putting an infringer on notice makes any infringement more likely to be seen as wilful, and therefore can enhance damages. But damages can accrue for infringement regardless. Waiting to sue until a good amount of damages has accrued is a technique often used by NPEs. – Maca – 2017-05-31T01:17:25.747

13

My answer is yes, and I agree with Maca. I'll add that your ability to enforce it might change in the future.

I'm not implying that your situation is the same as mine, but consider this:

In my industry at my company, we basically lack the ability to enforce the patent as well, but the more patents we have that our very large competitors infringe or might infringe, the less likely they are to try to sue us, either for any frivolous reason or for something we might be inadvertently infringing. I'm not sure, but I'd think this is fairly common in crowded technology areas.

In other words, we are basically unable to enforce, but also unable to easily withstand suits initiated by others as well and a patent portfolio is a reasonably priced arsenal for dissuading others from litigating against us due to counter-suit possibilities.

It's very similar to cold-war nuclear game theory.

Pete P

Posted 2017-05-24T08:15:56.290

Reputation: 131

2That's another very good point - I can confirm that mutual infringement is often tolerated in fear of costful lawsuits that would end as a lose - lose. – DonQuiKong – 2017-05-24T15:46:59.983

@Pete P This is actually bad for a little guy like me, if you think about it. To go back to your metaphor of nuclear game theory, I don't have any nukes, so unless I can form an alliance with one of the superpowers, any one of them will bully me into submission – DonkeyBoy – 2017-05-25T06:55:37.007

1Mutually Assured Destruction using patents? That's both awesome and sad at the same time. – Mast – 2017-05-25T08:54:30.347

@Mast - https://pbs.twimg.com/media/DAglMLDXkAE1RUa.jpg

– Eric – 2017-05-25T15:35:36.753

1@DonkeyBoy - the point is, if you have the patent, then they know that if they start to bully you, you can go find a superpower to cozy up to then. It gives you at least a little leverage, when before you had none, and that means that bullying you probably isn't worth it. – Ben Barden – 2017-05-25T15:41:21.000

@DonkeyBoy - I never worked in litigation, but I would expect that having even just one nuke would be a significantly better position than having no nukes. My suspicion is that any counterclaim (regardless of the number of alleged infringements) changes the landscape, from the big guy's perspective. – jdpatent – 2017-05-30T23:13:16.147

3

To my knowledge, depending where you are located, patenting something as abstract as an Algorithm is not even possible (EU). How to tie your laces is an Algorithm. Can you patent it? Hopefully not!

Gewure

Posted 2017-05-24T08:15:56.290

Reputation: 147

2If you look at my original question, you will see that I am specifically talking about computer code. That is not abstract. – DonkeyBoy – 2017-05-25T04:44:49.623

1@DonkeyBoy Computer code is not much different than just writing it down in clear prose. It is only creating a concrete representation of the algorithm, and anybody knowing the algorithm could do that. So that's nothing you can protect. – Volker Siegel – 2017-05-25T05:34:46.360

2@DonkeyBoy The usual "workaround" to convert an algorithm into something concrete enough to be patented is to combine it with some piece of hardware it interacts with. – Volker Siegel – 2017-05-25T05:36:16.877

@Volker Siegel Computer code that solves a scientific problem in a novel way is able to be patented and can be protected. I know this because I have seen many other patents in the same field as mine. – DonkeyBoy – 2017-05-25T06:50:28.723

@DonkeyBoy Yes, that is (my subjectiv view!) the terrible and absurd US-patent law, where you even can patent (like apple did) something as silly as an electronic information device with rounded edges. If you just define the design detailed enough. In the EU such patents are worthless/not even patentable. – Gewure – 2017-05-25T08:19:32.737

3@Gewure afaik the patent you are talking about was actually a design patent, which is specifically made to patent designs (not only in the US). – DonQuiKong – 2017-05-25T09:47:06.217

@Gewure Everyone is entitled to their subjective point of view, but you should consider why you have that point of view. Do you believe something just because the media or some other authority figure had that point of view. Or have you actually thought about this in detail and considered the implications? Are you really repulsed by the idea of inventors and designers from being able to protect their ideas and make a living doing what they love? And why? If I work hard for 5-10 years on a great idea to benefit society, do you disagree with my right to protect my hard work? – DonkeyBoy – 2017-05-25T13:33:30.697

This answer is simply wrong anyway: I don't see why it is upvoted. First, on a purely formal level, there's no EU patent (yet). Second, algorithms are perfectly patentable if they have a technical effect. Third, "abstract" is a US-centric term, which still does not exclude algorithms necessarily. – Maca – 2017-05-29T10:06:13.110

@Maca: there is no 'purely formal' law, yes. At is with most directives in the EU. In practice its still applicable: http://m.epo.org/index.html Depending on what algorithm you want to patent, your changes of getting it are significantly lower than in the US (where you can indeed patent anything). This is also the reason why there are many more patentlawsuits filed in the US than in the EU: you simply can't patent everything in the EU thats easily possible in the US.

– Gewure – 2017-05-29T10:43:35.710

@Gewure My point is the EPO has nothing to do with the EU. It's entirely independent. Turkey and Switzerland being members of the EPO is a clear signal of that. This is not merely a pedantic point, it is quite important: each country in the EU has its own patent laws which may be significantly different from those of the EPO. The UK (which remains an EU member at present) is the most notable example: UK patentability is wildly different from EPO patentability. This variation would not exist if the EU set patent law. – Maca – 2017-05-29T11:18:25.433

3@Gewure And the reason there are more patent lawsuits in the US in almost entirely to do with legal system, market size and expected returns. Treble damages plus jury awards in a market of 300 million people is much more lucrative than separate suits in the UK, France and Germany with judicially-set damages. That the US has historically had lower examination standards is relevant, but barely, since patentability is much more throughly contested during litigation. – Maca – 2017-05-29T11:21:28.837

1@Gewure And finally, at the risk of being long-winded, to say that anything is patentable in the US is to ignore the last decade of tightening standards, particularly post-Alice. – Maca – 2017-05-29T11:23:47.047

Ad Alice: "Relying on Mayo v. Prometheus, the court found that an abstract idea could not be patented just because it is implemented on a computer. In Alice, a software implementation of an escrow arrangement was not patent eligible because it is an implementation of an abstract idea."

  • back to the topic: An abstract Algorithm (Maths!) is not patentable, an implementation is patentable. This is why e.g. all graph-Algos (Kruskal, Djikstra...)are not patent-protected.

[..] – Gewure – 2017-05-29T12:12:54.093

So @Maca, no im not "simply wrong". Especially not in the EU. I don't know why you guys think you can patent a series of abstract steps?! This is anything but clear!

See e.g. https://en.wikipedia.org/wiki/Shoelace_formula , which derives its name from tieing shoe laces. I'll go and patent it now and make a lot of money ;)

– Gewure – 2017-05-29T12:13:47.110

Let us continue this discussion in chat.

– Gewure – 2017-05-30T07:46:30.183

1

I was in the same situation and wrote and submitted utility patent application myself. Unfortunately after "Alice" US supreme court decision virtually any algorithm could be qualified as"abstract idea", which happened in my case. With lawyer's help I rewrote the application to increase the chances and now waiting for a response from the USPTO. I requested the application unpublished so if it doesn't go through I can keep a trade secret. Good luck to you.

Pol99

Posted 2017-05-24T08:15:56.290

Reputation: 545

0

As it seems, you are assuming (I guess it is because it's typical on your industry) that someone will try to use your ideas without your permission. I believe that having them patented will, at least, give your the chance to fight (or to find a sponsor for a fight).

As I see your problem, if you file a patent, you benefits may derive from:

a) Selling your patent to someone who might make profit from it and has the resources to protect its legal rights on the patent

b) Commercial exploitation via royalties for licenses of the patent

c) Suing people who infringed your legal rights by using your patent without your permission (alone or with a sponsor)

On the other hand, if you don't patent your work, you will have no access to any of these benefits.

In addition, the worst thing that may happen to you is that someone steals your work, right? So, in that case, what other mechanism may help you to fight back better than a recognition of your ownership of the intellectual property with a patent.

Have you already thought of any other way to protect your work? Keeping it hidden doesn't seems like a very good plan, either...

Alexos

Posted 2017-05-24T08:15:56.290

Reputation: 1

-1

Let's say you created a patent, and big company A infringes it. Damn you, big company A. Cease and desist. Big company A laughs at you and says fuck you, you're tiny, we can do whatever we want. At this point, you turn around and contact your friends at big company B and say hey, I can sell you a cheap way to fuck over big company A. Big company B shakes your hand and buys you drinks and says you're our boy, you sell them your patent, and big company B sues big company A, making big company A sorry that they said fuck you to you in the first place.

So it's complicated, because patents can change hands. Transactions and future technology development can change the value of a patent, sometimes dramatically. Don't assume that just because you personally can't enforce your patent, no one will want to. A patent has no value if it is never written in the first place. Protect the idea first and foremost.

Oh, and in case no one has told you this, you can't patent an algorithm. Algorithms are math. You can't patent math. You can however patent a method or a VLSI or an apparatus, which is what everyone does instead. Go read a textbook on patent authorship.

johnwbyrd

Posted 2017-05-24T08:15:56.290

Reputation: 239