Thank you. As Peter mentioned, I'm an attorney for Google. We basically work on compliance issues around bringing open source into Google as well as releasing.

We're also working on a casebook of legal issues around open source. If anyone's interested in that, please take a look. Full disclosure, I'm not an engineer. I am an attorney. I practice in the US. So this is going to be based on US law, and it's a legal talk. So if that's not what you're expecting, you can go ahead and escape now before I get started.

Before I start, are any of you attorneys? OK. So I can make up a little bit, I guess. Are any of you familiar with patent law? What about copyright law? Probably copyright law. OK. I'm going to talk about patents today. I know it's not usual.

Most people think about copyright law when they think about open source, because open source is automatically protected by copyright as soon as you type that code. But we do have patent protection in open source. And some people might wonder, well, why would you want patent protection when you have copyright? So I'll just run through what they are. And again, this is based on US law.

So copyright, you get that the moment you create an expressive work. So the moment you write something, you do source code, whatever, you have that copyright. And if I asked all of you to describe this room today, you'd all have copyright in that description, even if two of you, for some reason, came up with the same exact description.

That's not the case for patents. Patents, first of all, you have to apply for it. It takes a long time. It's really expensive. And you can only get a patent for certain things. So in the US, it has to be new, original, and useful. So if I asked you all to come up with an idea and two of you came up with the same idea, you would both not be

able to get a patent on it. And it has to be something that isn't done before. Whereas if you write source code and you come up with something from scratch that's already been done, you'll still get the copyright for it. Copyright's the right to use and distribute. And patent is actually an exclusionary right. You have the right to prevent other people from making, selling, using your invention.

Copyright's going to last through your lifetime, plus 60 years, whereas a patent's only valid for 20 years. And then it goes in the public domain. And usually people are going to get a patent because they want to profit. So if you think about the pharmaceutical company, they have 20 years to recoup the cost of the drugs that they've spent years developing.

And that's why there's no other generic drugs on the market for those 20 years. Copyright's going to be protected in most countries, whereas patent is a territorial right. Unfortunately, you have to apply for it in every country that you want that right. And both of these have this thing called exhaustion, which I'm going to talk about in a little bit. So you can see why somebody might want both copyright and

patent protection. And if we think about that in the open source world, let's say I developed a code, which I cannot do, but let's say I did, to improve Google Maps. And I also have the patent to Google Maps. And let's say I gave all of you that code under a license to use for free.

This could pose a little dilemma. I'm giving you that code to use, but maybe by using it you're going to infringe my patent. So what does that mean? Can I sue you? You probably haven't heard of this because it's not really an issue in open source community, because the community has come to either accept open source licenses that explicitly mention patents.

So for example, the Apache 2.0 license has an explicit patent grant in there. It says, any patents that might be infringed by using the code I'm giving you, well, you have rights to that patent. So there's no catch-22. But in the spirit of open source, the community has also come to find some licenses that don't expressly mention

patents to also give you an implicit patent right. So if you think about the MIT license, it's super short, but it does give you the permission to use copy modify. So the community said, well, if you're giving somebody the permission to use that code, and by using that code they would infringe, that's not right, and that's not in the spirit of open source.

So that person is also getting an implicit patent grant. Now all this is great, there's no problems, but I want to think about source code for a minute. So if you notice the title is called Self-Replicating Technology. So copyright is really interesting because it can replicate itself, right? You can create a program that's going to make copies of itself as part of that program.

For example, there's some good computer viruses out there that will self-replicate themselves to be able to find bad computer viruses. And I think it's really interesting to see how the courts might address this because when patent laws were enacted, source code wasn't a thing. The courts have really struggled to keep up with

technology, especially in the US, and how to classify that technology. So just as an example, if you look at how the US courts have just looked at software patents. So I've tried to color code this, but let's go back to 1968. In 1968, the US Patent Office basically gave guidelines that said computer programs are unpatentable no matter how

you claim it. You cannot patent a computer program. In 1972, the US Supreme Court followed in line of that guidance, and they said an algorithm to convert binary coded decimal numbers into trust binary numbers is unpatentable. And they basically said that because they looked at it as math.

In the US, you can't patent math. You cannot patent E equals MC squared. But by the time we get to 1981, the courts are starting to catch up in having to think about this a lot more. And the Supreme Court actually ordered the Patent Office to grant a patent on an invention even though it utilized computer software.

And again, I'm giving you very basic. These cases are really complicated, but this is kind of what the Supreme Court and the court system is looking at. In 1998, the court said that software that yields a useful concrete intangible result can be considered patentable. Don't ask me what that means.

A lot of these cases cause a lot of confusion, and people never know whether they can patent their invention or not based on the court's guidance. But it looks like the courts are saying you can patent computer software. In 2008, that's in a green box because they didn't directly address computer claims. But the court's now of this machine or transformation test.

They said, well, if you use a machine or it transforms something, then your software can be patented. And if you think about it, at that time, a lot of people were like, well, I'll just slap a computer on my software, and I should be able to get a patent. In 2010, the Supreme Court started to get a little bit of cold feet, and we have this last case where they go

back to this idea, well, if this software is an abstract idea, can you do it in your head? Is it some sort of methodology that doesn't require anything tangible? Well, then you can't patent it. And we're all left confused. So we get to this big famous case. So if any of you were patent lawyers, you would know this case.

It's huge. It gave us the Alice test, which is just a big headache. So in this case, the patented invention was a computerized trading platform used for conducting financial transactions, software. And the court said, well, let's come up with a test to test all of these types of patents and claims. Step one are the claims at issue directed to a patent

ineligible abstract idea. That's something that cannot be patented in the US. Well, let's say it is. They're like, well, you can still salvage it. Is there an inventive concept or some element or combination of elements sufficient to ensure that the claim in practice amounts to significantly more than a

patent on an ineligible concept? We still don't know what this means. So in this case, they had added a computer to their software. And so the Supreme Court said, that's not enough. And they did not grant the patent. So now people can't just add a computer to their software and get a software patent. It has to pass this test.

And courts have, so in the US, you have the Supreme Court, which is the high court. And then you have lower courts. And a lot of the lower courts are really struggling with what does this mean. So that was really just to show you that software is in uncharted territory in case law. And it's hard for the law, which moves really slowly, to keep up.

And there's an interesting doctrine in patent law called patent exhaustion. They thought it would be fun to look at what does this mean in the sense of self-replicating technology. So what is patent exhaustion? Patent exhaustion in the US basically prevents the patent holder from profiting more than once on one item. So I sell you an item, I exhaust my patent right.

So if I sell my item to Peter, Peter can't be sued for patent infringement because he got it from me. But if he goes and sells it to, what's your name? Seth. Seth also isn't going to be sued for patent infringement. Not because he got the patent from me, but because I already exhausted my right when I sold it to Peter.

Makes sense, but this was derived back when patents were given for apparatuses, devices, machines, things like that. So what does that mean in the sense of software? Especially software that maybe can replicate itself. Is there a patent for every copy of that software that is made? So I want to look at some of the case law in the US

and see how that might apply. And before I jump in, Americans have tried to get around this for decades. They have come to the court system and said, well, what about this? And what about this? Can I still have my patent right? And the courts have been pretty clear about limiting that IP right. It's a one item, one reward type of doctrine.

So what if it's free? And this is very applicable in open source, right? Because open source is free. So the courts addressed this in a case called LifeScan Scotland v Shasta Texts. So LifeScan had a patent for a method of using blood glucose measuring devices. So essentially they gave out this meter

and then they had these little test strips that would be used in the meter. They gave out the meter for free or like very below cost to customers with the expectation that the customers would use their own test strips in the meters. And Shasta came along and started making their own test strips to use with the meter.

So LifeScan sued them and said, you're contributorily infringing our patent by making these test strips. So LifeScan sued them and Shasta was like, no, no, patent exhaustion, your patent right exhausted when you sold or gave that meter away. LifeScan said, well, we gave that for free.

Like the patent system is designed to give me a reward. If I'm giving that out for free, how can I exhaust my patent right? And so the court said, you can't evade it by just giving it away for free. And if you think about that, even in open source, there is an economic benefit for somebody using your software for free. I work for Google, there's a big economic benefit for us giving the search engine away for free, right?

So the court said, you can't get around it in that way. So then someone came and said, well, it's a method patent, right? So the exhaustion doctrine is usually applied to patents that have covered devices or physical items.

What about a method that's a strategy or like a series of steps? How could you possibly exhaust that? The court dealt with this in Quanta Computer v. LG Electronics. So in this case, LG licensed its patents to Intel and it licensed them to Intel saying, when you sell these patents to your customers,

you have to sell your customers that they can only use those patents with Intel products. So Intel did that, they told their customers that, but the customers didn't listen. And so one of those Quanta would combine the patents with other types of products like Dell, HP, Gateway, et cetera.

And so naturally LG sued Quanta for patent infringement. Quanta came back and said, patent exhaustion, you sold that to us, your right is gone. And the argument came up, well, it's a method patent, like how can you exhaust that? And the court said, you could. They said, a patented method may not be sold

in the same way as an article or device, but the methods can be embodied in some sort of product. And when that product is sold, you exhaust your patent rates. So then someone got really clever and said, well, can I contract around it? So lawyers like to write all sorts of contracts and try to get you a certain clauses.

So what if I can convince somebody to buy my patented item but not exhaust that right? So they would agree that the patent rate is not exhausted, is that possible? So the court had this question as well in the case, impression products v. Lexmark.

So Lexmark makes printers and ink cartridges. So if any of you have bought a printer, you know that it's the biggest pain to have to continuously buy ink cartridges, and that's what's the most expensive part. So Lexmark had two different situations for its customers. And the normal course customer can buy a cartridge at a regular price, use it, and then reuse it,

transfer it, there are companies out there that will refill the ink, no problem. Lexmark also had a program where it would sell its cartridges at a discounted price to its customers. And in that case, the customer could only use it and then throw it out. They couldn't reuse it, they couldn't sell it, transfer. But some customers naturally still do that.

And impression products was a company that would basically buy those ink cartridges and refill them. So Lexmark comes and sues them and says, you're contributorily infringing our patent by buying these from those types of customers. And of course, impression products says,

no, no, no, patent exhaustion. And Lexmark actually had a express patent restriction in its clause to its customers, saying that you can't do this, the patent right remains for those ink cartridges. And the court shot that down. The court said, when you sell an item under an express restriction,

you still can't retain patent rights. So maybe you have some sort of remedy in contract law, but not in patent law. And patent law is where the money is at. So you still can't get patent rights in that product. So we can see the court's trends so far. And so I'm gonna end with this last case, which is what if it's self replicating? And this is the case that kind of spurred the talk.

And in this case, it's called Bowman v. Monsanto. So it's actually, oh, thank you. It's actually a biotech case. So I'm sure you've heard of Monsanto. They make seeds. And so Monsanto has these Roundup ready soybeans that it will sell to its farmers. The farmers can use it once in a season.

And when they harvest the seeds that have been made from that, they can't replant them. Their option is to sell it to a grain elevator, and then it's used for consumption. So Bowman being the great farmer that he is, did that. But what he also did was he bought seeds from those grain elevators that had this Roundup ready soybeans in them,

and his farmers were selling them back, and he harvested those. He actually did it for like 11 cycles before Monsanto found out. And then Monsanto sued for patent infringement. And in this case, based on everything the court said, you would think, well, patent exhaustion, you've exhausted your right. But the court actually found that there was no patent exhaustion,

that the patent right was for the first seed and it wasn't for the seeds that had been made by that seed. And the court basically limited it to this very specific situation though. And it said, in another case, the article's self-replication might occur outside the purchaser's control. So it really honed in on the fact that Bowman had to buy those seeds and replant them.

It's not like he planted the seed and the seed itself kept planting itself. And then the court said, or it might be a necessary but incidental step in using the item for another purpose, and specifically talks about copyright. It is not a copyright infringement for the owner of a copy of a computer program to make another copy or adaptation,

provided that such a new copy is created as an essential step. So you think about the question I posed earlier, what if there's a program that has to create more copies in order for it to be used? Well, on the face of this case, it might seem that they have a patent right in each of those copies, but it sounds like the court would not apply the analysis on that case

because the program itself would be required to make those copies and you wouldn't come in and have to make another copy. If you made another copy yourself, then you could be liable for patent infringement. But in this case, the program has to make itself. So what does that mean for open source? Nothing terribly exciting except for the fact that I think the courts are starting to catch up

and I think they're starting to look at copyright a little more closely and understand what it really means and how it works. And I think in terms of open source, the patent protection is there and I don't think it's gonna change even when we start having programs like this. With that, I'll take any questions, comments.

Thank you, Sanal. We are a little bit ahead of time, so time for more questions, maybe. Thank you for your presentation. I want to ask,

are you aware about open invention network efforts regarding the patents around open source and the cases around it? And I don't know, what is your opinion about that effort?

So to be honest, I'm not completely familiar with it, but if you go to our case book, we actually do have a section in it. I love when I can promote this. We have a section in our case book. There's a patents chapter and there's a part on that. So if you wanted to take a look at that, you can get our thoughts on it.

And then there's some interesting discussion questions that you can think about as you read the cases. Yeah, thanks for your talk. I'm just wondering about data structures and proprietary formats. Is that patented or is it copyrights?

So if it's a functionality, if it's the way that you might do something, it's going to be patented. So if you're talking about source code straight up, it can be both. If you're talking about data structure in terms of like,

I'm not entirely sure, I'm an engineer, but. The file or an image type, or is it in GIS? There's kind of proprietary image formats, but I'm not sure if it's a patent or copyright. So if the image format is being displayed through like binary or source code, and that's what's displaying it, the binary source code, it can be both, but as you could see,

like the Supreme Court's not entirely sure about software patents, but it's definitely copyright. Anytime you write code, that's copyrighted immediately. But if you're trying to patent like the image or what it ultimately looks like, that would be patented. So you can get to the same image in multiple ways, but any of those ways would be patented because ultimately you are displaying this image,

if that makes sense. Is it? Okay. All right. Hopefully that was not too much legal talk. Well, I'll just say, as the prerogative of the chair,

what convinced you to come to Phosphor G as a Pacific place to talk about this? Is it because there's an interest in the geo side of this or because it's open source? I know this is being recorded, but I'll be completely honest. I saw a false and I thought open source. And we like to travel to different places

to just talk about open source and the legal side of it. And it wasn't until my talk was accepted that I realized what the 4G stood for. But it's been really interesting. I'm also, I do work a little bit on standards. So I've seen that come in throughout some of the talks in the conference. And I think just in general, open source,

at least Google's very committed to open source. And we try to encourage all Googlers to contribute. We try to make sure that we are being good stewardesses of our open source. And so that was kind of the main reason. And then also a little bit of sharing of patents, which is my passion and copyright and educating people so that they know whether their inventions

and copyright, copyrighted code are being protected and things like that. And also it's Romania, I've never been. It's a really good reason. Any final questions or comments? We're certainly delighted to have you here to talk about these issues.

Okay, so we've a few minutes to get ahead of the crowd to get over for lunch. So thanks again for your attention to session and thanks to the speakers.