Executive Interviews: Bruce Sewell The Role of Intellectual Property in Standards

"Intellectual property is an important component of any standard. It can be used to shape the standard and ensure that people can understand the risks and costs of participating. That's important if you want the standard to be broadly applicable and widely adopted."
—Bruce Sewell

Why is intellectual property (IP) important to Intel, and what is the role of IP in standards?

The products that Intel creates are the tangible embodiment of the intellectual effort that we put into our corporation. Intellectual property can be viewed as the legal embodiment of that effort. So from our perspective, IP is the legal corollary to products. The two are inseparable. Companies need both in order to do business.

Intellectual property is an important component of any standard. It can be used to shape the standard and ensure that people can understand the risks and costs of participating. That's important if you want the standard to be broadly applicable and widely adopted.

Don't open standards, which Intel actively promotes, require companies to give up their IP rights?

Open standards are sometimes confused with open or free access to IP, but in fact those are quite different concepts. From our perspective, an open standard has more to do with the process by which the standard is developed, and ensuring that people have access to the technology that's embodied in the standard. Whether or not you pay a licensing fee for the IP embodied in the standard is a separate issue. That's an important distinction.

There are some in our industry who advocate royalty-free access to the IP in standards. What is Intel's view on this? Is this a good approach or not?

It depends on the situation. There may be some cases where royalty-free could be appropriate, but to adopt that as a default�to suggest that in all cases it's better to be royalty free�we think is unfair to IP holders. It can also make it harder to develop a successful standard. To create a good standard, you want to get as many stakeholders involved as possible�including IP owners and those who will eventually adopt the standard. Forcing intellectual property owners to give up their IP in order to participate may make them stay outside of the standards bodies. So, a royalty-free policy could actually make it riskier to adopt the standard.

We believe it's more effective to choose a model, such as reasonable and nondiscriminatory (RAND) licensing, with clearly stated royalties and a restriction against discrimination. Then, people who want to participate in the standard have the absolute right to do so. There may be a license fee involved, but it's much more likely that IP holders will participate. That decreases the IP risk for adopters and increases the chances of a more robust and more widely adopted standard.

"We believe that a royalty�free approach to accessing the intellectual property in standards is unfair to IP holders and may make it harder to develop a successful standard. To create a good standard, you want to get as many stakeholders involved as possible. Forcing IP owners to give up their IP in order to participate will make them more likely to stay outside of standards bodies."
—Bruce Sewell

What if you decide not to participate? What are the risks of building a proprietary technology, compared to the risks of adopting the standard?

There may be legitimate reasons for a company to stay outside of a standard-setting activity. This may arise when the scope of the standard is too broad, or when the contribution being asked of one company�in terms of dollars, IP or both�is disproportionate to the contribution being made by the other participants.

Of course, the risk of staying outside the standard is that you don't remove the IP hurdles. If Intel were to stand outside of a particular standard, for example, we would keep our patents clear, but we could also be subject to the IP held by the companies that did participate.

On the other hand, there are some risks if you participate in the standard. For one thing, in most cases, you would be obligated to license your IP in order to participate. That could later compromise your ability to defend yourself against a challenge by an IP owner against one or more of your other technologies. Another risk is that the standard might have hidden pitfalls�patent risks or unanticipated licensing fees that are not clearly spelled out, for example. There's still another risk, if the standard and the proprietary technology end up competing in the marketplace. This can lead to a standards war, and these are generally not helpful for anyone, particularly end users.

So there could be risks whether you decide to participate or not. As a standards adopter or as an IP holder—and Intel is usually both—you have to balance the risks and benefits associated with going it alone versus the risks and benefits of joining the standard.

Of course, a key benefit of building a standard-compliant product is that it can interoperate with other products, and that may increase its overall market. So even though the obligation to license your IP could impair your ability to defend against a later IP challenge, the benefit of a larger market might outweigh that risk.

So you give up a little but get a whole lot back.

Absolutely. And that's particularly true if it's a good standard.

What constitutes a "good" standard, from Intel's perspective?

In our view, a standard is good if it is robust and widely adopted. We believe that a standard has a better chance of being robust and widely adopted if the major stakeholders participate in its development. We also believe that major stakeholders are more likely to participate if the development process is transparent, the costs associated with the standard are well understood, and the patent risks are well explained. It's not necessary that there be "no" risks. What's critical is that the risks are clear, so that you can adopt the standard with a greater degree of security. And that means the standard is more likely to become widely adopted.

Unfortunately, not all standards meet these criteria. For example, sometimes IP policies covering standards are vague, so the licensing obligations are not clear. Or the policies may dictate that the obligation to license IP extends to future iterations of the standard.

That's problematic for companies that want to participate. It means they can't be sure of which IP they may be obligated to license in the future, so there's no way to assess the risk of agreeing to license that IP. So clarity in the IP policy governing a standard is important.

Also, as I suggested earlier, we believe that a good standard, in most cases, is one that provides for reasonable and nondiscriminatory access to IP rather than royalty-free access. That's one of the issues being debated today—whether "open" should mean "free."

"In Intel's view, a standard is 'good' if it is robust and widely adopted. We believe that a standard has a better chance of being robust and widely adopted if the major stakeholders participate in developing the standard. We also believe that major stakeholders are more likely to participate if the development process is transparent, the costs associated with the standard are well understood, and the patent risks are well explained. It's not necessary that there be no risks. What's critical is that the risks are clear, so that you can adopt the standard with a greater degree of security."
– Bruce Sewell

Are there any other IP issues facing standards bodies today?

Patent searches are another critical area of concern. There has been some confusion in the law about which patents the participant in a standards development organization is obligated to license. A body of law that developed recently challenges whether participants in standards bodies have an affirmative duty to search out IP that might be covered by the standard, to disclose the IP and to affirmatively state that it reads on the standard.

If your portfolio consists of five patents, it is a reasonable exercise to do a search and determine that three of the patents may read on the standard and two of them do not. But if your portfolio consists of 5,000 patents, and you're not sure how the standard is going to evolve during the development process, an IP search becomes a massive, if not impossible, undertaking. Moreover, that undertaking requires both an engineering and a legal conclusion which could result in liability, if they are wrong. Many, if not all, major IP holders are going to view this as too risky and refuse to participate in the standards development body.

Where does the debate on patent searches stand?

At the moment it's being handled by the standards-setting bodies themselves. One of the problems that led to the legal challenge is that the language in standards agreements in the past was ambiguous as to whether there was an affirmative duty to disclose. That's been resolved with much clearer language in standards agreements. Now many standards bodies are incorporating phrases such as "reasonable knowledge" or "reasonable apprehension" in their agreements�terms that mean, in a legal sense, that your obligation doesn't extend beyond what you know, that you don't have an affirmative duty to search. We believe this is an effective approach to the problem.

What other global challenges do standards bodies face today?

Globalization itself is a key challenge, and a reason why standards are becoming even more important to Intel as a company and to our industry as a whole. Now companies are moving into countries that may have developed very different ways of doing business. To interact with those countries, we all need to develop standards that allow both sides to speak the same language from a technical standpoint.

I think that Intel is well positioned to help address the evolving legal and technology challenges posed by globalization. We have a pretty broad portfolio of technologies, and some fairly sophisticated people working on standards issues. And we try to be a leader, by sharing our knowledge, tackling problems facing standards bodies, and finding ways to work with our industry colleagues, so that collectively we can become more effective in the standards arena.

About D. Bruce Sewell

Bruce Sewell is senior vice president and general counsel of Intel Corporation. He joined Intel in 1995. From 1996 to 1999, Sewell was the group counsel for the Enterprise Platforms Group. In 1999, he took over the team of attorneys, contract managers and paralegals who provided legal services to the Intel Architecture Group. Sewell also advises Intel on complex transactions, such as technology licenses, alliances and acquisitions.

Technology