The DOJ Understands Royalty-Free Licensing is Far from Free

“Royalty-free” is an appealing notion, especially when applied to the licensing of a patent essential to a standard. This standard is even more compelling to an implementer when every participant in the relevant standards body or proprietary consortia, including preeminent technology companies, has pledged it will grant its essential patents on a royalty-free (RF) basis.

What could be better than an entire ecosystem around a standard or proprietary solution and its implementation being bound by an RF pledge? This notion – universal royalty-free reciprocity – is the long-standing model of the W3C Consortium. The model has been adopted by the leading developer of a video codec, the Alliance for Open Media (AOM) – a proprietary consortium.

Is it the case that “royalty-free” applied to standard-essential patents (SEPs) is FRAND (fair, reasonable and nondiscriminatory), but better? Today’s U.S. Department of Justice (DOJ) doesn’t think so.

According to Dina Kallay, Deputy Assistant Attorney General in the Antitrust Division of the DOJ, the “royalty-free” claim raises concerns. In a recent speech, Kallay called out  “private consortia that impose mandatory, royalty-free cross-licensing obligations on their members. . . . [I]n the context of a proprietary consortium that is made up of dominant implementers that collectively possess market power, such arrangements can be competitively harmful.” She argued that “royalty-free”, coupled with forced reciprocity, “can effectively operate as [a] collusive scheme[] among dominant players to promote closed, proprietary standards.”

‘Reciprocity’

There’s a lot more to reciprocity than meets the eye. The IP rights (IPR) policies of both W3C and AOM are particularly dense, but each distinguishes three levels of reciprocity.  In both, the term covers multiple tiers of patent licensing:

• The first, basic reciprocity, is fairly uncontroversial. It covers the relationship between an SEP holder licensing to another and the licensor’s pledge of a RF license granted in exchange for the licensee’s matching commitment.
• The second, licensee’s RF pledge to all others, is already problematic: under the AOM Patent License 1.0 the licensee is obligated to “make its [SEPs] available under this License . . . [for any] Implementation”.
• The third tier pushes the RF framework to universal RF reciprocity: the licensee, and thereafter its own licensees (and the licensees of those licensees) “must reproduce this License with any Implementation”.

How does an implementer of AOM’s AV1 standard show its assent to the terms of the AOM Patent License 1.0?  It needn’t bother because it is bound when it “exercises patent rights granted under this License”. The AV1 implementer doesn’t even have to click through to join the ecosystem.

When ‘Free’ Means ‘Give Me Everything You Have’

There are several noteworthy concerns that arise under universal RF reciprocity. The first, and most obvious, is that “royalty-free” has different economic values among SEP holders. For two holders of AV1 patents essential to the specifications accepting RF licensing, the value of foregone royalties can be wholly disproportionate: if X has one AV1 essential patent and Y has 100, the price Y pays for a licence of X’s patent, when Y becomes subject to the RF terms, is the value of the royalties Y would have otherwise received. This is not limited to royalties otherwise payable by X. Instead, Y is forgoing payment of the royalties by implementers throughout the entire AV1 ecosystem.

AOM has developed the AV1 specifications in a technology – video codecs – where standards bodies have traditionally adopted FRAND policies (that is, required licensing terms that are fair, reasonable and nondiscriminatory, including payment of reasonable royalties). For an implementer (a non-AOM participant) that also has a portfolio of AV1-essential patents, the concern with the AOM policy is clear: the price it pays for the RF patents to be licensed by the AOM participant is the royalties it would have otherwise received not only from the AOM participant but also from any other AV1 implementer.

‘Free’ Impedes Participation by Innovators 

A second concern is the effect on innovation in standards creation. Companies at the forefront of technological development are often leaders in innovation and in the manufacture of devices that incorporate this innovation. They should be able to seek returns not only from sales of their devices but also from royalty payments from other manufacturers. If the price of their participation in standards development means that they must accept an IPR policy calling for universal RF reciprocity, then it is more likely they will refuse to join, and their innovative solutions will not be brought to the standards body.

Technological progress through standardization will be hamstrung if these innovators are absent. Alternatively, if they choose to remain, innovation will be stifled if the innovators know that, for every discovery, the return can be captured by universal RF reciprocity.

‘Free’ Creates an Opaque Patent Landscape

A third concern is that the notion of universal RF reciprocity inverts the licensing landscape. For standardized technologies, the IPR model overwhelmingly adopted by standard development organizations (SDOs) is FRAND coupled with transparency. During development of a standard, the SDO participant contributes its patented innovation, while declaring its patent potentially essential to that standard and affirming that a license covering that patent is available on FRAND terms. Transparency and FRAND terms are also the twin goals of regulators.

Universal RF reciprocity guts this process: there is no need to give a FRAND affirmation because all the patents are licensable for $0; while there is no need to identify the patents or the patent owners because, arguably, in a RF ecosystem, declarations would only provide information – infringed patents, names of patent holders – ultimately unnecessary for implementation.  In the RF world, implementers do not seek out the owners of the patents they infringe, and patent owners do not care about infringers because all they can expect is nothing. From this position, there are at least two consequences.

First, the murkiness of the SEP landscape makes it difficult to determine what IP is held, not only by participants in the work of the SDO or proprietary consortium, but also by those who are not members. The SDO or proprietary consortium and its members may engage in sharp practices; for example, intentionally including innovations from non-participants with the hope that they won’t find out or that in the end they’ll accept RF licensing. Should SDO/proprietary consortium participants have substantial market power, they may be able to impose the technology and force its quick adoption. Non-participant innovators would feel pressured into adopting the technology and accepting the RF reciprocity without even having the opportunity to know which patents will be neutered as a consequence. After all, a due diligence of their entire portfolio against the standard/proprietary specifications is a prohibitively expensive task in terms of time and financial resources. This reality would force them to accept RF reciprocity without having been able to make an informed decision.

A second consequence to this flouting of FRAND rules and SDO/proprietary consortium transparency is that the SEP licensing process is discredited. When an SEP holder contacts an implementer to negotiate licensing terms for a purported RF standard, the implementer’s first response could well be disbelief:  how can the SEP holder expect a royalty on products implementing a standard/proprietary specifications widely proclaimed to be royalty-free? The disparagement of common licensing practices (such as negotiating a royalty with a patentee) will only increase the risk of hold-out.

From the point of view of implementers, the RF community skews the market for standard-compliant devices. Even if some bilateral negotiations are feasible (resulting in royalty-bearing licenses), licensees are hobbled by a royalty burden that their RF competitors don’t share. They are obliged, for example, to offer products, identical to their RF competitors, at a higher price. For an assessment under FRAND rules, “royalty-free” fails the nondiscriminatory prong and results in an anticompetitive impact in the product market.

‘Free’, But with Exceptions

Finally, can there really be a complete, all-encompassing RF ecosystem?  Has, for example, AOM achieved its objective and, using its doctrine of universal RF reciprocity, convinced, cajoled or coerced all AV1 SEP holders silently to acquiesce in forswearing patent royalties?  The answer is “No”:  SEP holders, who neither participated in AOM specification development nor implemented AV1, and thus are not subject to the AOM IPR policy, have formed an AV1 pool of royalty-bearing patents. Moreover, at least one AV1 patent holder has, consistent with W3C rules, opted out of RF licensing.

So, is AOM deceiving the market by overstating its claim that AV1 is royalty-free? While not expressly calling out AOM, Kallay notes the risk of a questionable claim of universal RF reciprocity, characterizing it as “touting the benefits of . . . royalty-free standards”.

Department of Justice Assessment of Universal Royalty-Free Reciprocity

In her remarks devoted to proprietary standards development policies that are not FRAND based, Kallay noted that “pooling or cross-licensing arrangements can have pro-competitive characteristics”. She went on to present the case of a proprietary consortium “that imposes mandatory, royalty-free cross-licensing, [a policy that] allows a group of dominant implementers to fix the price of royalties at zero”. Based on the mechanism of universal RF reciprocity, “companies that need to use the standard are forced to license their patented technologies for free.” Kallay concluded:

“In such cases, these royalty-free cross-licensing requirements can effectively operate as collusive schemes among dominant players to promote closed, proprietary standards [with] the effect of stifling innovation. . . . Over time, successful proprietary standards may push out other truly open standards that would better support interoperability and innovation.”

Another aspect highlighted by Kallay is that by characterizing the RF framework as FRAND, proponents could be making “a misrepresentation of material information to the marketplace . . . a cause for competitive concern”. The proprietary consortium authoring such standards may fail to meet the criteria for a standards development body under U.S. rules (e.g. OMB Circular A-119).

Kallay’s speech presages greater attention by U.S. antitrust authorities on anticompetitive conduct arising out of collective action by market dominant players within an SDO (or in the formation of the SDO and the adoption of its IPR policy). This focus, both in the United States and elsewhere, is a positive development that can protect competitive SEP licensing markets and enhance innovation rewarded by royalties.  It will also benefit new entrants, and small and medium-sized enterprises, including those that are solely research-oriented and those that can couple their R&D with manufacturing SEP implementing products.

Image Source: Deposit Photos
Author: iqoncept
Image ID: 61369901