Thursday, June 2, 2011

Lodsys's infringement theory could work against Apple (and Google)

Lodsys disagrees with Apple's position that its patent license covers app developers and has already sued the first seven app developers, accusing mostly iPhone and iPad apps, but also one Mac app and one Android app.

Before I look into the arguments Lodsys might use against Apple and (in connection with Android) against Google, let me explain why it's interesting but irrelevant until Apple and Google give blanket coverage to all of their app developers (the ones that have been sued and all those who could be sued any moment).

Without blanket coverage, app devs will lose

The primary concern at this point is whether or not Apple will give blanket coverage to app developers. If Apple or Google (depending on who makes the platform/s on which your app runs) doesn't, then all questions of law and fact in this case can be reduced to this one: does Lodsys have enough of a story that its case has a good chance of not being dismissed early on? And by "dismissed" I mean that it would be the end of the story, not the kind of dismissal that allows the refiling of an amended (for example, more specific) complaint.

In my opinion, Lodsys does have a case that's unfortunately good enough to (potentially) go to trial. The fewest app developers would financially survive that scenario (possibly not even one of the seven defendants named so far), so the only responsible friendly advice one could give is that all app developers approached by Lodsys should sign a license agreement provided that its terms don't make things even worse.

When I say that it may "go to trial", bear in mind that this case will be put before a jury in East Texas (unless someone successfully gets it transferred to another district, which is a long shot at this stage). That court is known to process these cases relatively swiftly (by comparison with most other U.S. district courts) and to hand a high percentage of patent holder-friendly decisions. By the way, the European equivalent of the Eastern District of Texas is the Landgericht (district court) of Düsseldorf, Germany, and Apple filed a lawsuit against Nokia with that court last fall. Almost half of all European patent litigation is put before that one court because patent holders (who instigate most patent-related lawsuits) -- including Apple -- love it so much.

Getting back to the U. S. District Court for the Eastern District of Texas, it's not only known for judges who tend to be on the patent holders' side but also juries that are easily persuaded by the idea that lone "inventors" (such as Mr. Abelow, whose patents are now asserted by Lodsys) should be protected against "infringers". The ability of those juries to understand technical issues (such as the API issues at the core of Apple's letter to Lodsys) is often criticized. Red Hat CEO Jim Whitehurst claims that in that court you "generally [find] a jury that has not completed college." That statement is too condescending for my taste, but to give you an idea, in the dispute that resulted in RIM's famous $612.5 million settlement with NTP, the closest thing to a patent expert on the entire jury was a high school teacher.

So assuming you as a defendant aren't already bankrupt before the jury convenes, you have to assume that the judges in that district will brief the jury in a patent holder-friendly way and Lodsys's lawyers would tell them the story of how you steal Mr. Abelow's intellectual property. Against that background, the jury may hand a verdict against you, possibly awarding millions of dollars in damages. And then there may still be lawyers who say that the jury was wrong and you should have the decision overturned, but then you'll have to appeal to the Court of Appeals for the Federal Circuit, which you may not be able to afford.

So it all comes down to what Apple and/or Google -- depending on the platform(s) for which you develop -- will do for you. If they don't protect you 100%, then even 99% is not enough and you should seriously consider paying Lodsys. But let's hope and -- for the time being -- assume that the platform maker(s) will cover you. In that case, or if you're simply curious about the intricacies of such a patent dispute, it's interesting to take a look at how Lodsys's argument might play out against Apple's defense, and why Apple's defense is far from certain to work -- and why Apple's approach would be much less likely to work for Google in connection with Android...

Defenses against patent infringement claims

If large companies with vast legal resources get sued over patents, you will see them present a long list of affirmative defenses. They will, figurately speaking, throw in the kitchen sink.

This not an exhaustive list of defenses, but typically, deep-pocketed defendants will

  • present several reasons why they believe the patent shouldn't have been granted (including but not limited to prior art),

  • claim that (even if the patent was considered valid) there is no infringement because their accused technologies don't practice the claimed invention,

  • and in many cases come up with theories as to why even if the patent was valid and infringed, they would have a right to do what they do, or the patent holder wouldn't have the right to enforce it against them (misuse, estoppel etc.).

That was an oversimplified description, and I purposely used a lot of non-legalese, plain-English terms to convey the idea.

Compared to that kitchen-sink approach, Apple's letter to Lodsys was very narrow: without contesting the validity of the patent or the infringement theory, Apple's argument is essentially just that Apple's license to Lodsys's patents also takes care of app developers (in the context in which Lodsys appears to assert its patents against app developers so far). Apple didn't waive its right to bring up other defenses, but it would have been good if Apple had attacked Lodsys more broadly.

It's now important that Google also take a position on this issue, given that Android app developers have received letters from Lodsys and that one of the accused products in Lodsys's app dev lawsuit is an Android game. I believe that if and when Google does speak out, it will probably also use Apple's this-is-licensed-stuff argument (although, as I will explain further below, that argument is harder to make for Google than for Apple) but additionally raise other defenses. Google was very creative in that respect in its dispute with Oracle (on which I regularly report, most recently when Google hired a third law firm for its defense).

Divided infringement: not a convincing defense

As soon as Lodsys's assertion letters became known, patent professionals with an interest in this (particularly Apple fanbois) started to discuss possible defenses. One particularly popular theory was that of "divided infringement". If Apple becomes formally involved with this litigation, it may try to make that point as well -- affirmative defenses are allowed to contradict each other, and this theory would run counter to Apple's license-based exhaustion theory.

The idea of divided infringement is that a patent is not violated by a single product or service, but by a collective of two or more independent actors, none of whom can be held responsible. Lodsys listed four patents in its letters to app devs, sued over two of them, and had previously provided claim charts (a claim chart is a document describing an infringement theory) for each accused app in connection with one of them (the '078 patent). That patent describes a "system" consisting of different elements, and one could try to develop a theory of divided infringement based on the fact that some of the technology in question is provided or (in the case of Apple's central app store server) operated by Apple while some of it belongs to the app, claiming that neither Apple's stuff nor the app infringes the patent on its own to a material extent -- it takes two to tango, and, therefore, no single one should be liable for infringement.

The relevant CAFC opinion was handed in the BMC Resources v. Paymentech case, and it contains an important limitation:

A party cannot avoid infringement, however, simply by contracting out steps of a patented process to another entity. In those cases, the party in control would be liable for direct infringement. It would be unfair indeed for the mastermind in such situations to escape liability.

I said before that divided infringement is infringement of the kind that "takes two to tango". The above limitation relates to situations in which there are two who tango, but one of them has hired and directs (like a choreographer) the other. In that case, the CAFC said, the one who's in charge -- the "mastermind", as the CAFC calls him or her -- must of course be liable. (This may remind Starcraft fans of the Zerg Overmind.)

If an app dev accused by Lodsys of patent infringement used the divided infringement defense, Lodsys could overcome that hurdle by convincing the court that the app dev is the "mastermind" because an accused app is in charge: it provides "control or direction" because it makes calls to Apple's API that result in actions which, in Lodsys's opinion, infringe the '078 patent. This includes API calls that result in the storage of information on Apple's central servers.

I'm skeptical of the divided infringement theory. I think the apps are in charge. While an in-app purchasing process includes various details that aren't micromanaged by the app, claim 1 of the '078 patent is a very high-level description of a process, and at that high level, I believe the app must probably be deemed to be in control.

Let's look at the individual elements of the patented "system" from the perspective of control:

  • "units of a commodity that can be used by respective users in different locations": according to a claim chart I saw, it's not clear to me whether Lodsys describes each installation of an app as a unit of a commodity, or whether Lodsys means portable devices on which the app runs; in the first case the app devs are solely responsible for this component of the claimed system; alternatively, Lodsys would argue that an app running on a device controls the device as far as the alleged infringement is concerned

  • "a user interface" that elicits information from users: while an operating system will provide user interface controls and the underlying program code, it's again a component that is really controlled by the app (the operating system only draws and manages the user interfaces in accordance with the directions provided by the app via API calls)

  • a "memory [...] capable of storing results of the [...] interaction": in a physical sense that memory is part of the device, but in a logical and more relevant sense, we're talking about the app simply having set aside a small portion of the machine's memory (most likely through the declaration of one or more variables) to store information; so again, the app is in charge

  • a "communication element [...] capable of carrying [information elicited from the user] from each [device] to a central location": the apps don't have that capability all by themselves, but by calling the related communications functions of their operating system's API, the apps definitely control and direct this data transfer

  • a "component capable of managing the interactions of the users in different locations and collecting the results of the interactions at the central location": while I don't believe in a divided infringement theory in connection with the previous bullet points, this item raises a more interesting question because here we're talking about servers, not just the client-side hardware and software that the app controls locally; but considering the language of the Paymentech decision, Lodsys could argue that sending messages across a wireless network to a central server is also a process that is controlled and directed by the app; the argument would be that the operation and provision of the central server is basically contracted out to Apple, Google or whoever runs it.

The exhaustion argument

Now let's look at Apple's theory: the exhaustion argument.

It's very difficult to form an opinion on the merits of this kind of argument without having access to the relevant license agreement (which in all likelihood is an agreement into which Apple entered with Intellectual Ventures a while ago). The outcome of any decision on this defense could be determined by the specific language in that license agreement. Apparently, Lodsys believes to be in a strong position and would even like Apple to publish Lodsys's position on this, but Lodsys itself can't do so because of confidentiality obligations. Considering that many app devs are concerned, it would be great if Apple could be transparent about this and publish Lodsys's letter (unless Apple has confidentiality obligations with other parties, such as Intellectual Ventures, that make this impossible).

But even without knowing the exact language of the license agreement, it's apparent that Lodsys has a whole lot of confidence in its position and one can look at the known facts and compare them to the applicable cases based on certain, somewhat realistic, assumptions.

The idea of "exhaustion" is that if a patent holder grants a license to one licensee, that licensee's customers shouldn't need to take a license of their own for using the licensee's product in a way that the original license already covers. It's like, "you don't get paid twice for the same thing."

The first Supreme Court decision on this goes back to the year 1873 (Adams v. Burke), and there were subsequent decisions in 1917 (Motion Picture Patents Co. v. Universal Film Mfg. Co.) and 1942 (United States v. Univis). These days -- and especially in connection with information technology products -- the 2008 decision in Quanta Computer v. LG Electronics is the most frequently cited case. It is also referenced by Apple in its letter to Lodsys (just before the closing paragraph).

To apply that ruling to the Lodsys dispute, if Apple's own technology "substantially embodies" the '078 patent in connection with what Lodsys claims are infringements by app devs, Apple's app devs are fine, but the aforementioned substantial embodiment depends on whether Apple's own technology has "no reasonable noninfringing use" and includes "all the inventive aspects of the patented methods."

Yes, this is the critical part, and it's where it gets most complicated: we must also look at the claimed "system" as a whole and ask ourselves

  • whether Apple or Google provides technology that "substantially embodies" that claimed invention (anything less than "substantially" embodying it makes this defense fail),

    which depends on

    • whether there's no reasonable noninfringing use of Apple's or Google's technology (if there is reasonable noninfringing use, this defense fails), and

    • whether Apple's or Google's technology includes "all the inventive aspects of the patented methods" (if any inventive aspect is missing, this defense fails).

Note that the criterion of "no reasonable nonfringing use" always has to be seen in light of the asserted patent and the context in which it is asserted (in-app upgrades). There are definitely ways to use iPhones, iPads or Android devices that have nothing to do with Lodsys's '078 patent -- such as playing a computer game locally that never sends data to a central server. The question is whether any reasonable implementation of Apple's or Google's in-app purchasing API will inevitably result in something on which Lodsys's '078 patent reads.

In my opinion, there's a reasonably strong case for inevitability. I can't imagine any reasonable way to implement an in-app upgrade without having a user interface that asks the user about it. Without asking the user, there would be no contractual basis for charging -- one could only do free automatic upgrades (like some web browsers do nowadays, but they don't charge). Lodsys might try to argue that free upgrades about which the user isn't even asked constitute a possible form of noninfringing use, but that wouldn't be a strong point.

What I'm less convinced of is that the platform maker's technology indeed includes "all the inventive aspects of the patented methods". It doesn't have to include non-inventive aspects (such as "the application of common processes or the addition of standard parts"). However, all of the inventive stuff has to be there.

My personal opinion is that the '078 patent is not at all inventive: neither as a whole nor in any of its parts. But unfortunately, the USPTO determined that it was inventive (otherwise the patent wouldn't have been granted), and there's a significant risk that a court (not only but particularly in East Texas) will also apply a standard for inventiveness that is much lower than the one I'd like to see applied.

The lower that inventiveness standard is, the more "inventive aspects" there are, and with every inventive aspect increases the risk of something essential not being provided by Apple or Google. If Lodsys convinces the court that at least one inventive aspect is missing, app devs will (if other criteria are also met) be liable.

Let's look at potentially inventive aspects on an item-by-item basis:

  • "units of a commodity that can be used by respective users in different locations": Apple has a strong point that it supplies such commodities; but most Android-based devices are not sold by Google itself and Lodsys could try to have the term "commodity" construed in a way that would work against Google.

  • "a user interface, which is part of each of the units of the commodity, configured to provide a medium for two-way local interaction between one of the users and the corresponding unit of the commodity, and further configured to elicit, from a user, information about the user's perception of the commodity": while any mobile operating system contains the underlying technology that's needed to put such a user interface together (user interface controls that are drawn and code to process any of the related events), the tricky part here is that those UI elements are of a generic nature and Lodsys might argue that its patent describes something that's more specific than what the mobile operating systems provide.

  • "a memory within each of the units of the commodity capable of storing results of the two-way local interaction, the results including elicited information about user perception of the commodity": Apple can try to argue (as it does in its letter to Lodsys) that it provides "memory". In a physical sense, Apple does. Google doesn't in connection with Android-based devices made by third parties and could at best argue that its operating system and/or virtual machine provides memory management. But even for Apple (and even more so for Google) it's a problem that Lodsys's claims are more specific and relate to memory in which a certain kind of information is stored (such as variables defined with the appropriate data types). So again, if what Lodsys describes in its patent is considered inventive (contrary to my personal assessment), there's a potential problem here.

  • "a communication element associated with each of the units of the commodity capable of carrying results of the two-way local interaction from each of the units of the commodity to a central location": Apple provides the hardware and software for this and would have a strong case concerning this component. With Android-based devices not supplied by Google there could be a debate over whether supplying the operating system is sufficient or not.

  • "a component capable of managing the interactions of the users in different locations and collecting the results of the interactions at the central location": here we're talking about the server side, and both Apple and Google would have a strong point in connection with this element.

I know that all of this is terribly abstract and complicated, and software developers like to write code rather than worry about any of that. Unfortunately, the law of the land is that anyone, even a little app developer, can be sued for software patent infringement. I just tried to point out that there's considerable risk in the Lodsys dispute.

Apple is a powerful company. So is Google, which I hope will also come out in support of app devs as soon as possible. But even powerful companies don't make the law.

Ultimately, the matters of fact in a case like this will be put before a jury on which the closest thing to a technical expert might be a high school teacher (like in NTP v. RIM), or on which there might not even be one person who completed college (according to Red Hat's CEO's view of East Texas). That jury won't have the last word if you appeal successfully, but that's a big "if".

Given that there is so much uncertainty, I urge all app devs to focus on what the platform makers really guarantee to you. No position taken by their legal departments -- no matter how forcefully it is pronounced -- will ever be a substitute for the coverage you need if you take the risk of embarking on a court battle with someone like Lodsys.

If you'd like to be updated on the smartphone patent disputes and other intellectual property matters I cover, please subscribe to my RSS feed (in the right-hand column) and/or follow me on Twitter @FOSSpatents.

Share with other professionals via LinkedIn: