Have you discovered The Alexandria Project?
The acronym “FRAND” is very much in the news today, and with good reason. The battle to control, or at least share in the bounty, of the mobile marketplace has motivated technology leviathans like Google, Samsung, Microsoft and Apple to bring every tool and weapon to the fore in order to avoid being left in the dust. So intense is the competition that not only standards, but the finer details relating to the pledging of patents to facilitate the implementation of standards, have become the subject of headlines in the technology press.
The purpose of this blog post is not to report on the skirmishing that is still ongoing, but to peel off and explain the multiple layers of nuance and tactical opportunity that underlie the seemingly simple concept of “FRAND.” How many layers? Let’s just say that you may lose count before we cover everything you need to know to make sense out of what is really going on behind the scenes.
Isn’t it time you read:
The Alexandria Project?
a Tale of Treachery and Technology
Remarkably accurate while consistently spellbinding: I ran across a reference to this book at a blog unrelated to the author, and after reading one chapter, bought the book
Great thriller: In the spirit of Vincent Flynn and Tom Clancy, this cyber-security thriller is a great read. Compelling characters, great detail and an an unsettlingly plausible scenario add up to a real page-turner.
Delightfully unpredictable! Updegrove has managed what many attempt but few can execute: a plot that is both credible and surprising….A great read – I can’t wait for the next one!
Strong characters and compelling plot: I read a lot of novels and this is a very good one. The characters are believable and engaging and the plot is compelling with several clever twists along the way….Highly recommended
Excellent and accessible techno-thriller: Updegrove…clearly knows the subject matter inside and out, but is too self-assured and smooth a writer to hide behind that insider’s knowledge….I look forward to Updegrove’s next book with great anticipation.
Great Read: This is a very well written, highly engaging story. The scary thing about it is that the entire plot is far too possible to come to life.
Fantastic! The Alexandria Project is a gripping novel of intrigue and suspense. The characters may be fictional, but we all know their real-life equivalents. The storyline may be fiction – but maybe not.
Read these and more 5 Star reviews at Amazon
The only part that’s fictional is that it hasn’t happened yet
Available Now for $2.99 or less
Buy at Amazon
Buy at iTunes Store
Buy at Barnes & Noble
Given the complexity of the subject, I’ll revert to a device I picked up from Stephen O’Grady at Redmonk – the self-interview – because I’ve found it to be a very useful way to cover material that is not only complicated, but which also needs to be related to specific events in the news. So here we go.
Q: What does “FRAND” mean?
A: FRAND stands for “fair, reasonable and non-discriminatory,” and must be followed by more words to complete the intention (usually, those words are “licensing terms”). It’s an umbrella phrase that doesn’t designate specific terms. Instead, it expresses the bottom line result that a license must add up to.
Q: Does “RAND” mean the same thing?
A: Yes. FRAND is the normative term in Europe, while RAND predominates in the U.S., where we’re a bit more impatient. And query what the difference is between “fair” and “reasonable,” anyway?
Q: In what context are these terms used?
A: Standards setting organizations (SSOs) require those that help create a standard (and sometimes all of the members of the SSO) to state before a standard is approved for implementation whether they have any patent claims that would be unavoidably infringed by someone implementing the standard (such claims are referred to as “Necessary” or “Essential” Claims in SSO intellectual property rights (IPR) policies). If a member does, or may, have such claims, they are required to state which of up to three elections it will make in connection with those claims (whether all three alternatives are permitted depends on the SSO):
– they will license their Necessary Claims without charge, and otherwise on FRAND terms, to implementers of the standard (they can also agree to simply not assert their Necessary Claims against implementers)
– they make the same commitment, but reserve the right to charge a royalty or other fee, which must be reasonable in amount
– they refuse to license one or more Necessary Claims, and identify those claims, and the portions of the draft standard that would result in the infringement.
The extra requirements in the third alternative are quite important, because they allow the working group to try and "design around" the withheld claims. Some SSOs require those choosing the second alternative to identify the claims, and the effected portion of the standard, as well, so that the working group members can decide whether to try to design their way around the claims in order to avoid the chance that implementers may have to pay a royalty.
Q: Fair enough. But what does this really mean?
A: Ah, now there’s an interesting question. First, let’s look a little more closely at what the constituent words in these acronyms mean at a high level.
First, there’s “Fair” and “Reasonable.” It’s obvious that words like these at least mean “not exploitive or extortionate.” But that gets us only so far down the road.
It’s also obvious that these are comparative words, so by nature, they imply that there is some sort of benchmark in relation to which fairness and reasonableness can be measured. But what would that benchmark be?
Interestingly enough, there really isn’t one. In fact, the parties to a lot of recent litigation, between individual companies (like Broadcom and Qualcomm) and between regulators and companies (like the European Commission and the U.S. Federal Trade Commission (FTC), on the one hand, and Rambus, on the other hand) have grappled at length with this issue.
And it’s trickier than first meets the eye, as the FTC found out to its sorrow when it tried to punish Rambus. By definition, a standard grants a monopoly power to the owner of a Necessary Claim under that standard. So should fair and reasonable be determined by reference to a situation where there are multiple ways to build something – a commodity situation – or where there is only one way – a monopoly situation?
Another dynamic that makes the pricing terms really tricky is that an implementer will often need to license other patent claims from the same company in order to build their product, and the patent owner will price those claims along with the Necessary Claims for a single, bundled price. So how can you tell how much is allocable to the Necessary Claims, and how much to the other claims?
Suffice it to say that while there has been an increasing amount of litigation on this subject, there hasn’t yet been enough to generate a clear legal definition upon which the marketplace can rely.
Q: Okay, I can understand that. So how about “non-discriminatory?”
A: At first blush, that would seem to be an easier phrase to apply. In essence, what it means is that I won’t license my Necessary Claim to one vendor for $1 a unit built to a standard, and for $2 to another vendor for the same purpose.
But not so fast. What if one vendor wants to buy build 100 units a year that conform to the standard, and another wants to build a million? Shouldn’t the latter be entitled to a quantity discount?
Here’s another complication: what if the owner of the Necessary Claim already has a cross license in place with the first vender, under which both licensor and licensee already have access to each others’ patents? Is it non-discriminatory for the owner of the Necessary Claim to not charge the first vendor anything per unit, and charge the second $1 per sale? The parties to the cross license would say yes, because each invested the money to develop the technology to begin with, and then to patent it. So each is getting value from the other already.
But if it is fair, then how will new, emerging technology companies be able to enter the marketplace with standards-compliant products, when they have to pay royalties to multiple huge, established, incumbent companies, while no money changes hands between those behemoths?
So, just as a “most favored nation” clause in a traditional contract often doesn’t help the party that demands the term, because contracts can differ one from another in so many ways, the requirement to provide “non-discriminatory” terms to all implementers is at best hard to define, and at worst, provides little protection to standards implementers.
Q: So does “FRAND” mean anything specific at all?
A: Yes, there are some terms that are generally assumed to be FRAND terms. Sometimes they’re expressly mentioned in an SSO’s IPR Policy and sometimes not. An example is “defensive suspension,” which means that if you license a Necessary Claim to an implementer, and they turn around and sue you for infringing a Necessary Claim that they own when you implement the same standard, then you can withdraw your license. But it wouldn’t be permissible to withdraw your license if they sued you for some other reason.
Q: Is there anything an implementer, and especially a small company, can do if it thinks that it hasn’t been offered FRAND terms?
A: Almost nothing, for two reasons.
First, Necessary Claims licenses often include confidentiality terms. So one company has no way to tell whether it got the same terms as another company – even another company in a similar situation. So that makes violation of the “non-discriminatory” part impossible to tell.
An implementer requesting a license will know, of course, how much they’re being asked to pay, and if that seems like too much, they could refuse to pay. But as noted above, there isn’t a clear legal definition of what a “reasonable” payment rate would be. So if the implementer thinks the amount demanded is too high, it would have to call the owner of the Essential Claim’s bluff, and see whether they sue.
Of course, if the implementer does get sued, they’re going to have to either pay up, or pay to defend themselves and hope that a judge or jury agrees with them that the amount demanded was not reasonable. But even if the implementer wins, they may still have to pay a lower royalty – in other words, the amount that the court determined would be reasonable.
As a result, there are more suits that involved an implementer claiming either that the claim wasn’t infringed at all, or that the patent claim itself should never have been granted by the Patent and Trademark Office. If can succeed on one or these arguments, then it won’t owe anything a all.
Unfortunately for the implementer, if neither of those avenues look promising, then there’s not much else they can do. The SSOs aren’t any help, and for good reason. They simply don’t have the resources to get involved in disputes between members. Also, the antitrust rules (in the U.S.), competition laws (in Europe) and similar rules elsewhere in the world make it almost impossible for SSOs to get involved in pricing discussions. And remember that most information and communications technology SSOs are membership organizations, so there aren’t a lot of incentives for them to get involved when their members start throwing stones at each other.
Implementers that think they are being unfairly treated can, however, contact regulators to complain, and sometimes those regulators will open an investigation to see what’s going on.
Q: So why don’t SSOs just define what words like “reasonable” and “non-discriminatory” mean? Wouldn’t that make life easier for everyone?
A: Well, that depends on what you mean by “easier.” Vendors, and especially vendors with very large patent portfolios, care greatly about maintaining their freedom of action. In one situation, they might want “reasonable” to mean one thing, while in another they may want it to mean something different.
For example, where a particular standard is crucial to a multi-billion dollar strategic decision by a vendor, they’d like FRAND to mean as tight a commitment as possible, so that all of their competitors have to make their Necessary Claims available, ideally on royalty-free terms. But in another situation, where the same vendor is more or less just going along with a particular standard for the ride, they might want to be able to use their Necessary Claim card to their advantage against one or more other companies.
And I expect that the last thing that companies with large patent portfolios would like to see would be a public discussion over whether cross-licenses between major companies should be counted the same as a requirement to pay a significant royalty by a company with few, or no, patents to its name. If a regulator were to take the position that a little company should get a free license in a case like that, then a major barrier to entry for new competitors would fall.
Q: But what about the litigation you mentioned? Doesn’t that cost companies a lot of money?
A: Yes and no. Yes, it costs those companies that engage in litigation a lot of money. But every year, while tens of thousands of companies give FRAND commitments, only a handful of companies (or fewer) end up in court. The rest either reach agreement, or the licensee was simply unable to tell whether the licensor was fulfilling its FRAND commitment or not.
Q: So why don’t the regulators get involved? Don’t they care?
A: They do care, but let’s take a closer look at what they care about.
In the U.S., the regulators have (in my view), a pretty balanced outlook. They want to protect the consumer against unfair practices by vendors and service providers, but they also don’t want to place gratuitous burdens on companies (this is, after all, not only the Valhalla of capitalists, but an economy that’s still struggling to emerge from a dire recession) so it’s unlikely that they are going to go overboard on overburdening industry to no purpose.
I meet with a variety of U.S. regulators every year, and they’re alert to issues that may be important for them to attend to. So far, no one is pointing at the interpretation and application of FRAND as a major issue, so to date they haven’t taken any overt actions, although their radar has been spinning.
In Europe, the situation is somewhat different. First, there are fewer major, multinational technology companies headquartered there, so they are more watchful to what foreign (often U.S.) companies are up to. But those same U.S. companies are also careful to put lots of plants, and therefore hire lots of people, in Europe. They employ lots of lobbyists there, too.
On the other hand, U.S. headquartered companies are constantly sending complaints to the European Commission, objecting about the conduct of other U.S. companies. Sometimes, they lodge these complaints directly, and sometimes they use trade associations as their proxies.
As a generality, the European Commission has been responsive to these proddings. In recent years, they’ve looked into standards-related activities of a number of large U.S. technology companies, including Microsoft, IBM, Google, and more. Sometimes they’ve levied penalties, and sometimes they have simply closed these inquiries without taking action. Now that Neelie Kroess is no longer at the helm, some of the pressure on multinational companies has been relieved.
At the end of the day, perhaps the best summary is that regulators have to cover an awful lot of issues, and their priorities sometimes change as their leaders change. Absent a clear crisis, there’s only so much attention they can pay to any individual issue, absent a clear indication of why they should expend significant public resources in that cause.
Q: Well, is there an indication that they should, in the case of FRAND abuse?
A: I hate to be tedious, but the answer again is “yes and no.” Let’s talk first about the “no.”
Despite the fact that (to my knowledge) not one, single SSO defines what “FRAND” means in any detail, and only a small (but growing) minority of all SSOs require members that participate in the development of a standard to provide royalty-free terms (thus eliminating most, but not all, of the ambiguity in a FRAND commitment), very, very few members ever actually assert a Necessary Claim against an implementer.
Mostly, this is the case in market sectors like ecommerce, Internet, Web and some software domains.
The “yes” part arises in other market sectors, like consumer electronics, where not only are FRAND assertions very numerous, but those assertions are coupled with royalty requirements that are frequently asserted in court if necessary to compel compliance.
Q: So why do some companies make their Necessary Claims available for free at all?
A: Sometimes it’s because a market for their products won’t exist unless the standard takes hold. Think of WiFi, for example. Unless lots of people implement the standard, you won’t know to ask for it, or be satisfied with the results. So making some patents available for free makes good business sense, in order to make it more likely that the standards in question are adopted more broadly and more quickly. This is even more the case where there is more than one standard competing for adoption, and you’ve bet the ranch that one standard will succeed instead of the other.
But patent owners can also make a lot of money from other patent claims when they are successful in steering a standard in their direction. Imagine that we’re creating the first standards for bicycle hand operated brakes, and you own patents on calipers that can squeeze against bicycle tire rims. You magnanimously offer to make those patents available not only on FRAND terms, but for free, and hand-operated caliper brakes win the day.
But guess what? You also own patents that cover the rubber pads that the calipers hold, and the standard doesn’t address pads at all. Now, anyone that wants to build hand-operated caliper brakes has to come to you for a license, and you don’t have any obligations at all to make these patents available on reasonable or non-discriminatory terms. Voila!
Q: That doesn’t sound good. But still, in those market sectors where Necessary Claims assertions are infrequent, why don’t SSOs just change their IPR Policies to require – what do you call them – FRAND-Free policies?
A: (Usually, they’re called FRAND-Z, where the Z stands for “zero,” as in “zero cost”).
Well, it’s interesting you should say this. Some major multinational companies, like IBM, are increasingly advocating for FRAND-Z policies.
And Microsoft, which has traditionally argued for FRAND as the default rule for SSO IPR Policies, last week said that it would not [added:] seek an injunction against sales of standards-compliant products where the implementer had not secured a FRAND license from Microsoft, and where
assert non-FRAND royalty rights in essential claims against standards implementers, even where Microsoft it had not helped develop those standards, or was otherwise bound by a FRAND obligation (e.g., where it had acquired a patent that was already subject to a FRAND obligation) , and therefore had no obligation to limit their demands. That was a very powerful statement, and quite a change of policy.
Q. So why did they do that?
A: Well, the proximate cause (as a scientist would say), was that Microsoft presumably wanted to make Google look bad. You see, Google is trying to acquire a key mobile company called Motorola Mobility, and European and U.S. regulators have to approve that acquisition before it can go through. Motorola Mobility owns a variety of patents, including patents with Necessary Claims, which apply to mobile devices. Google told the regulators that it would not charge an implementer more than 2.25% of a device’s net cost to use Google’s patents. And it also reserved the right to seek an injunction when an implementer refused to sign a license upon those terms.
Q: Well, that doesn’t sound like a lot. What’s the big deal?
A: How about if I told you that the same device would likely implement more than 1,000 standards, and that multiple companies would own Necessary Claims under many of those standards? If every patent owner took the same position, a mobile device would cost, what, over $10,000?
Q: I take your point. Still, if Google thought it could make that much off its patent claims and reserved the right to get an injunction to help enforce its intention, why would companies like Microsoft and Apple promise not to use the same tactics?
charge an excessive royalty for theirs?
A: Good for you – you’re on to the really significant part of Apple and Microsoft’s statements.
Presumably these two companies took a critically objective look in the metaphorical mirror, and realized that it simply would no longer fly for them to declare war over
charge for a Necessary Claim under an important standard anyway, so why not take the opportunity to poke Google in the nose, and make a public commitment not to? And that was huge.
Q: Why do you say “huge?”
A: Because it means that these enormous, incredibly influential, stockholder-focused companies realized that open standards, like open source software, are becoming so important to the marketplace that the damage they would do to their customer relations if they took advantage of the monopoly power handed to them by an SSO would far outweigh any profit they could hope to obtain from the extra licensing royalties.
It takes some historical perspective to appreciate how big a seismic change this represents. In the mid-1990s – in other words, back in the bad old days of proprietary platforms – “open standards” were something that everyone talked about, but no one really believed in.
After a decade of ever increasing reliance on the Internet, the Web, and now the Cloud, you simply can’t bull* about open standards anymore. Standards are what allow our brave, ever more exponentially inter-connected world to work. If you mess with standards, well, maybe you’re an enemy of the people.
In short, just like open source software, open content, and open data, more and more people are recognizing that open standards are part of what we might reasonably refer to as the “Web Commons.” All of us are more and more dependent on that Web Commons in order to survive and prosper. If the patents that allow the Internet and the Web to operate aren’t available on FRAND terms, what are we to do?
I think that the concept and scope of a Web Commons is beginning to take hold in the public consciousness. And unlike the old assumptions that the exploitation of any commons is inevitable (the so-called, tragedy of the commons), the new gestalt is that the Web Commons isn’t there to be exploited. More and more people are deciding that in the world of tomorrow, the Web Commons must be guarded with all of the globally distributed might and determination that humanity can muster. So anyone that wants to unfairly tax, or exploit, the Web Commons will do so at their great peril.
Q: That sounds pretty cool. So can you wrap this up?
A: I think so. What the events of the last few weeks indicate is that perhaps it’s no longer necessary to define what “FRAND” means. Like open source software, more and more people recognize the value and importance of open standards. And just like open source software, you don’t necessarily have to have a narrow definition of what FRAND means. You simply know it when you see it, not by its individual characteristics, but by whether or not it offends the community.
And just like open source, if you take action against one user of open standards, you will increasingly be seen as taking an action against all users of open standards – in other words, against all users of technology. If you decide to do so nevertheless, then you’d better be prepared to answer to that wider audience as well.
If you think about it, you might say that’s not only fair, but reasonable, too.