Joint Patent Infringement in Cloud Computing

Richard Gilly

Law Technology News

08-28-2012

At first blush, it may seem straightforward that to infringe a patent an accused infringer must make, use, sell, or import all elements of the patented system, or must perform each step of the patented method. But two cases under en banc review by the U.S. Court of Appeals for the Federal Circuit, and the series of recent decisions leading up to these two cases, highlight how complicated the question of infringement can become in internetworked and networked computer systems such as in public and private "clouds."

Computer or software patents often claim their inventions as computer processes, computer systems, or as computer programs. The two en banc cases awaiting decision by the circuit court, McKesson Technologies, Inc. v. Epic Systems Corp., 98 U.S.P.Q. 2d 1281 (Fed. Cir. 2011), en banc rehearing, No. 2010-1291, and Akamai Technologies v. Limelight Networks, 629 F.3d 1311 (Fed. Circ. 2010), en banc rehearing, No. 2009-1372, deal with the situation where the claimed invention is a computer process requiring multiple parties to be involved for the claim to be infringed, a situation known as joint or divided infringement. Under current case law, for such a method claim to be directly infringed, every step of the claimed method must be practiced. Canton Bio Medical v. Integrated Linear Technologies, 216 F.3d 1367, 1370 (Fed. Cir. 2000). In the context of a computer system, a party who does not practice all the steps of the claimed computer method cannot be found liable for direct infringement unless such party exercises "direction or control" over other parties who perform the missing steps on behalf of the first party. BMC Resources v. Paymentech, 498 F.3d 1373, 1379 (Fed. Cir. 2007).

The previous holdings in Akamai and McKesson (now under en banc review) had confirmed the above "direction or control" standard, also referred to as vicarious liability, and went further to hold that an agency relationship or contractual obligation to perform any remaining claim steps must also exist for a party to be vicariously liable for patent infringement. Akamai, 629 F.3d at 1320; McKesson, 98 U.S.P.Q. 2d at 1283. In McKesson, this requirement proved elusive, as the court found multiple parties performing the claimed method were unrelated, and so the patent holder failed to show a single entity that practiced all the claimed elements.

In Akamai, a software provider was accused of infringing a claim directed to a content delivery service for Web pages that included replicating page objects by tagging embedded objects on a page. Akamai, 629 F.3d at 1316. The tagging, however, was performed by customers of accused software provider Limelight, and not by Limelight itself. Accordingly, since the defendant in Akamai did not perform all of the claimed steps, the Federal Circuit affirmed the district court judgment of noninfringement.

The en banc court in McKesson and Akamai is now reconsidering whether direct infringement requires a contractual or agency relationship as part of the required direction or control, or under what other circumstances the actions of multiple, unrelated entities can directly infringe, individually or collectively.

Regardless of how these cases are decided en banc, liability for joint infringement will usually depend on how a claim has been drafted to try to capture the many and varied computing environments, multiple actors, and other distributive aspects inherent in 21st-century cloud computing. In other words, the frontline challenge to protect computer-related inventions and enforce rights to such inventions will depend on the foresight of the patent drafter. As such, the patent practitioner should work closely with inventors to imagine, anticipate, or otherwise envision the scenarios under which the computer-related invention could be implemented now and in the future.

PATENT CLAIMS TO COMPUTER SYSTEMS

Though inventions of computer-related patents are often claimed as computer-implemented methods, they may likewise be claimed as computer systems comprising various hardware and software elements, as well as computer program products themselves. The McKesson and Akamai cases discussed above are directed to computer process claims performed by multiple actors, and thus do not directly address how infringement of computer system or apparatus claims would be analyzed when the system is owned or controlled by different parties, such as in Centillion Data Systems v. Qwest Communications, 631 F.3d 1279 (Fed. Cir. 2011).

The Federal Circuit in Centillion analyzed an infringement action against a software and service provider, Qwest, brought by front-end customers operating their own personal computer systems. The court first addressed whether there was infringing use of the claimed system, such system being capable of infringement by making, using, or selling it. The court held that the software provider's customers, not Qwest itself, were infringing users because such customers put all the elements of the Qwest system into "service," such as by activating or otherwise making both the front-end and back-end elements work for their intended purposes, and obtaining a benefit therefrom. Centillion, 631 F.3d at 1284. In so holding, the court rejected a more stringent standard argued by Qwest: for customers to "use" the system, the customers must exercise physical or direct control over each system element. Qwest's control of certain back-end system components and operations would have resulted in the customer's use failing to meet this more stringent standard.

The Centillion holding relating to customers' use of a computer system may give computer system patent holders some leeway in establishing infringing use by end-users. But the case may be more significant for the hurdles it imposes on patent holders accusing the software providers themselves of direct infringement of a system claim. First, the Federal Circuit in Centillion held that Qwest could not at law be an infringing user of the claimed system because the presence of the independent personal computer users at the front end of the system meant that Qwest never put the claimed personal computer data processing means into "service." In other words, Qwest did not control the ultimate system nor obtain benefit therefrom. The court went on to state, "Supplying the software for the customer to use is not the same as using the system." Centillion, 631 F.3d at 1286.

The court held further that the accused software provider Qwest could not be considered a user unless it were vicariously liable for its customers' actions -- unless Qwest satisfied the same "direction and control" standard applicable to establishing vicarious liability in the method or process claims involving multiple users in Akamai and McKesson discussed above. The court ruled that the software vendor's providing software and technical assistance to its customers did not make it vicariously liable for its customers' actions, nor did the software vendor sufficiently direct its customers or use them as its agents. Centillion, 631 F.3d at 1287.

The patent holder in Centillion likewise did not prevail on its claim that the software provider infringed the patent by "making" an infringing system, because the software provider did not make the claimed "personal computer data processing means" or install the client software to complete the claimed system. The infringing act of selling, though not explicitly ruled upon by the Centillion court, may have likewise failed under similar reasoning.

INDUCEMENT OF INFRINGEMENT TO THE RESCUE?

As seen in the above cases, it may be the customers of the accused infringers who are the sole direct infringers by virtue of using a patented system or practicing all the steps of a patented method. For that reason, to reach infringement by a software provider, a patentee may need to establish that the software provider induced infringement of its customers or other end users. Such inducement of infringement is a type of indirect infringement, and has likewise been the subject of some recent caselaw which may pose new challenges to establishing inducement of infringement. First of all, inducement of infringement requires there to be an underlying direct infringement, Dynacore Holdings Corp. v. U.S. Philips, 363 F.3d 1263, 1272 (Fed. Cir. 2004), such as was found in Centillion for the customers. Even when end users are the accused direct infringers, a finding of direct infringement may prove elusive, as occurred in the McKesson case, discussed above. In Centillion, the customer was held to be a user who directly infringed, but the Federal Circuit did not have before it the question of whether the software provider, Qwest, induced such customers' direct infringement. There was thus no ruling on the issue of inducement of infringement.

In addition to establishing direct infringement, to prove inducement of infringement, a patent holder is required to show that the accused inducer had the requisite "intent" or "knowledge" to cause the direct infringement. DSU Medical v. JMS Co., 471 F.3d 1293 (Fed. Cir. 2006). A recent Supreme Court case, Global Tech v. SEB, S.A., 131 S.Ct. 2060 (2011), addressed this issue, but, in so doing, may have raised the bar on establishing the requisite intent required for inducement of infringement by software providers, at least in certain cases. By way of background, prior to the Global Tech holding discussed below, there were some diverging holdings as to how specific the accused inducer's "intent" needed to be to constitute inducement of infringement. Thus, for example, in some cases it was sufficient for the accused inducer to know that it was causing a third party to perform certain acts, without knowing whether such acts amounted to infringement. Hewlett Packard v. Bausch & Lomb, 909 F.2d 1464, 1469 (Fed. Cir. 1990). In other cases it was necessary to establish that the accused infringer knew (or should have known) not just of the acts performed, but also that the third parties' acts would constitute direct infringement. Manville Sales Corp. v. Paramount Systems, 917 F. 2d 544, 553 (Fed. Cir. 1990); DSU Medical, 471 F.3d at 1293.

Against this backdrop, the Supreme Court in the Global Tech case seems to have sided with the more stringent standard enunciated by DSU and Manville, holding that induced infringement requires that the party knew (or should have known) the induced acts constituted patent infringement. The Supreme Court also went on to hold that a patentee can establish this knowledge requirement by demonstrating that the accused inducer engaged in "willful blindness," a well-established doctrine that would be satisfied in the context of induced infringement if the defendant subjectively believes there is a high probability that a fact exists, and then takes deliberate actions to avoid learning of that fact. Global Tech, 131 S.Ct. at 2070.

Evidence establishing that an accused inducer of infringement possessed the requisite knowledge will, of course, remain fact specific. For this reason if the accused inducer has received a qualified opinion of noninfringement, such opinion may constitute evidence that the accused inducer did not know the induced acts would constitute patent infringement and thus did not have the requisite specific intent. See Broadcom Corp. v. Qualcomm, Inc., 543 F.3d 683, 699 (Fed. Cir. 2008). It should be noted, however, that the existence of such a noninfringement opinion, while it may constitute evidence, even compelling evidence, may likewise not be sufficient under some circumstances to dispose of the issue of inducement of infringement. See, e.g., mCube Corp. v. Sea Change, Int'l., 436 F.3d 1317, 1324-25 (Fed. Cir. 2006).

In view of the foregoing, establishing infringement of claims involving multiple actors may present not only the challenges of joint infringement, but also those related to establishing inducement of infringement. And successfully establishing liability for induced infringement poses the additional legal hurdles outlined above, especially those related to the requisite knowledge of the accused inducer now enunciated by the Supreme Court in Global Tech.

CONCLUSION

In patent infringement litigation, the name of the game still remains the claim, which is all the more so in computer system patents in the 21st century. Foresight in drafting patent claims to cover the many and varied computer architectures available both today and tomorrow will remain challenging and hindsight, unfortunately, is often 20/20. Patent attorneys and agents may wish to consider one or more "brainstorming" sessions with inventors or other technical personnel to make doubly sure that patent claims focus on as few actors as possible or as few system components as possible to be more readily enforceable down the road.

Using both system and method claims remains an advisable approach because, as seen from the above discussions, the two types of claims may be regarded differently by the courts, depending on the accused infringer or the particular computer architecture claimed.

In the area of cloud computing, it is possible that the users of patented systems are not necessarily individual customers, but could be larger companies and thus worthwhile targets for enforcement of patented computer inventions. For this reason, user-focused patent claims may remain a viable alternative for certain cloud-based computer service providers or systems.

Software providers may likewise wish to consider the value of obtaining opinions of noninfringement, as such opinions may constitute evidence that the software provider lacked the requisite intent or knowledge to induce infringement by others of an asserted patent.

The much-anticipated decisions by the Federal Circuit in Akamai and McKesson will likely clarify the standards for establishing infringement of a computer-based method in the multi-actor context. At the same time, however, it is likely that the en banc decisions will raise further questions that will only be answered as the Federal Circuit decisions are applied by lower courts in the future. For this reason it remains prudent to adopt a claim program covering the invention through different techniques, so that the claims may remain viable in the face of the vicissitudes of claiming computer-related inventions in the 21st century.

Richard Gilly is a shareholder in the intellectual property practice group at Akerman Senterfitt. He specializes in patent and trademark law, including litigation and prosecution matters, and has regularly served as the lead attorney and lead patent counsel in patent infringement litigations. E-mail: richard.gilly@akerman.com.

This article originally appeared in Law Technology News.