3 Functional Claiming Questions To Address Post-Williamson

July 6, 2015

(as published in IP Law360)

A patent claim is functional if it defines the claimed invention in terms of what it does rather than what it is. Over the last 20 years, the U.S. Court of Appeals for the Federal Circuit relaxed restrictions on the construction of functional claims — leading to a sharp increase in their number. The Federal Circuit’s recent decision in Williamson v. Citrix Online LLC calls these functional claims into question. Williamson signals a possible return to old restrictions on functional claiming and, in particular, a return to "means-plus-function" claiming: limiting a functional claim to what is disclosed by the specification, or equivalents. Such a return threatens to significantly narrow the scope, and impact the validity, of software patents that rely on functional claiming. Innovators and practitioners should therefore closely evaluate what Williamson means for their offensive and defensive patent strategy and should put a plan in place to deal with its potential implications.

The Rise of Functional Claiming

An element of a patent claim is functional if the element defines the claimed invention, or a feature of the invention, in terms of what it does (its function) rather than what it is (its structure). “Means plus function” claiming under 35 U.S.C. § 112(f) represents a bargain between Congress and innovators: Inventors may claim their inventions functionally, but if they do, the claim will only be interpreted to cover the “structure” (and reasonable equivalents) described in the patent’s text and drawings for performing that function. Where no examples of structure are given, a means-plus-function claim is invalid. In the case of software, the “structure” has been interpreted to be the detailed steps of the algorithm for performing the function, as well as the hardware on which the software executes.

Means-plus-function claiming has fallen into disfavor because the claims are typically narrower in scope and litigation over those claims is typically more complex. But this disfavor does not extend to functional claiming generally. Indeed, inventors prefer functional claims because — absent the means-plus-function “bargain” — they are typically broader than claims reciting detailed structure. On the back of Federal Circuit precedent specifically holding that functional claiming not subject to the means-plus-function bargain is permissible,[1] there has been a rise in functional claiming in computer technologies.

For example, such claims often take the form of a “computer,” “circuit” or “processor” “for performing” or “configured to perform” a given function. In the current, generally accepted view of such a claim, the computer, circuit or processor is a structural term. Until Williamson, Federal Circuit precedent[2] provided a “strong” presumption against construing a claim element as being in means-plus-function format where there was such a “structural” term used together with the function, rather than the more generic “means for” performing the function. Though some criticized this presumption as permitting inventors to circumvent the means-plus-function bargain, this precedent gave patent owners near certainty that by using terms such as “computer for,” “circuit for” or “processor for” instead of “means for,” functionally claimed computer inventions would enjoy broader scope than those subject to means-plus-function construction.

The Impact of Williamson

In the Williamson decision, the whole Federal Circuit (i.e., acting “en banc”) overturned this “strong” presumption that applied when the words “means for” were not used. Under the new standard, courts will still apply a presumption, but the presumption is no longer “strong.” Instead, courts will use a “balanced scale” to evaluate whether the claim itself recites what an ordinary skilled person would interpret as “sufficiently definite structure” for performing the claimed function.[3] If a challenger shows it does not recite such definite structure, the claim is subject to the means-plus-function rules. Applying this new analysis to the “distributed learning control module for [performing a function]” limitation at issue, the Federal Circuit found that the “module” did not recite sufficient structure and should be construed as a means-plus-function limitation.

The Federal Circuit’s restructure of the means-plus-function analysis in Williamson creates uncertainty for functional claiming. In Williamson’s aftermath, innovators and practitioners should evaluate three questions to determine Williamson’s impact on their patent strategy. First, what does Williamson mean for issued software or hardware patents claimed functionally? Second, how should software and hardware patents be claimed to avoid unintentional means-plus-function construction? Third, how does this decision impact strategies that software patent owners and challengers have been developing surrounding the new “abstractness” doctrine following last year’s Alice v. CLS Bank decision from the U.S. Supreme Court?

1. New Defenses, New Challenges: Williamson’s Impact on Issued Patents

Attorneys drafting software patents have now long relied on Federal Circuit precedent supporting that a functional claim that generically recites a “computer” or “circuit” that performs a function typically connotes sufficient structure to avoid the means-plus-function rules.[4] Williamson calls into question both this precedent and the multitude of claims relying on it.

This could drastically change the scope of a patent owner’s existing portfolio. For example, pre-Williamson decisions from the patent office’s appeals board have explicitly found that functional claims reciting structure such as “a processor in communication with the memory, ... programmed to,” “a processor ... configured with the program to” and “a processor adapted to” did not recite sufficient structure and consequently were means-plus-function limitations. This was even with the prior “strong” presumption against such a construction.[5] Post-Williamson, district courts might be inclined to follow the patent office’s lead.

From a defendant’s perspective, the removal of the “strong” presumption presents new potential defenses. When a functional claim is asserted, the patent’s text and drawings should be carefully reviewed to identify the examples of structure for that function. If the text and drawings disclose examples, and the functional claim is limited to those examples in line with the Williamson decision, there might be a noninfringement position or clear design-around to resolve the dispute quickly. If the text and drawings do not disclose examples, the claim may be invalid.

From a patent owner’s perspective, more claims subject to the means-plus-function rules make asserting patents more costly. Litigation resources will be expended construing the exact bounds of the otherwise clearly worded claims. Further, if such a claim term is construed under the means-plus-function rules, it is likely that more extensive discovery and source code review will be required to ensure that an infringing product performs the disclosed algorithm. This can make proving infringement and settlement discussions more difficult.

2. Renewed Focus On Structure: Williamson’s Impact on Claim Drafting

By removing the presumption against applicability of the means-plus-function rules absent “means for” in the claim, Williamson removes a bright line for claim drafters. Patent applicants and their attorneys now need to take extra care if they wish to avoid an inadvertent means-plus-function construction.

Renewed focus should be placed on describing implementation examples in the text and drawings, and on using language from those examples in the claim, to highlight the overlap between the “structure” described in the patent and recited in the claim. Claims that recite more explicitly how each function is performed may also convey sufficient structure to avoid the means-plus-function rules. Reciting the components of the computer and the software, including the boundaries of — and interaction between — the different hardware and software modules, may also be helpful.

Williamson also reinforces the need for a patent strategy discussion to be two-part: business and technical. On the business side, it is now more important to discuss the economic goals to be achieved through patenting the invention to ensure that a patent will be tailored to the owner’s business. This tailoring is particularly critical with respect to claim scope, as the Williamson decision and other recent decisions show that very broad functional claims will be difficult to obtain and difficult to assert. Costs may be significantly reduced by ensuring that claims are narrowly focused on supporting the client’s business goals.

On the technical side, attorneys should therefore take care to ensure that inventors are given enough protection to safeguard their idea, but that claims are not drafted so broadly that the structure of the invention is lost. Often this starts in the invention disclosure phase of prosecution, where variants on the invention may be discussed, developed, and incorporated into the specification. This allows the attorney to both distill these variants into the essence of the invention, as well as draft claims of varying scope to separate features of the invention, to ensure the client’s business is well protected.

3. Different Problem, Same Solutions: Williamson and the Abstractness Doctrine

Commentators have noted since Bilski that the rise of the abstractness doctrine is due to the rise of functional claiming producing overbroad claims. With this in mind, it is unsurprising that the two-part test set out by the Supreme Court in Alice for determining abstractness has parallels to determining whether a functional limitation should be construed as means plus function.

In Alice, the first step for determining whether a claim is abstract is to identify whether there is an abstract idea.[6] In Williamson, the first step for evaluating whether a claim is means plus function is to identify whether the claim is functional.[7] In Alice, the second step is to determine whether the claim adds “something more” to the abstract idea.[8] In Williamson, the second step is to determine whether the words of the claim recite enough structure for performing the function to avoid a means-plus-function construction.[9]

This parallel can be seen most recently in Internet Patents Corp. v. Active Network Inc., decided June 23, 2015.[10] In that case, the Federal Circuit panel found functional claims invalid under the abstractness doctrine. Importantly, the Internet Patents court noted — in support of finding that the claim did not add “something more” to the abstract idea under step two of Alice — that the claim “contains no restriction on how [the function] is accomplished” other than a generic computer.[11] In other words, the fact that the claim was functional weighed against its validity under § 101.

This parallel means that Williamson reinforces the lessons taught by Alice for those seeking patent protection: (1) Be more explicit in your disclosure and (2) Be sure to have claims broad enough to protect the invention, but not so broad as to render them abstract.

For those defending against suit, this parallel creates an additional avenue to defend against functional claims. Presenting both means-plus-function and §101 defenses can put the plaintiff in a difficult position. If the plaintiff argues for the broader claim scope and the claim itself provides no limitation on how the function is accomplished other than a generic computer, it risks invalidity under the abstractness doctrine. If the plaintiff argues for the narrower scope conveyed by the means-plus-function limitation, it risks a finding of noninfringement or indefiniteness, depending on the disclosure in the specification.

[1] Halliburton Energy Servs. Inc. v. MI LLC, 514 F.3d 1244, 1255-56 (Fed. Cir. 2008).

[2] Lighting World Inc. v. Birchwood Lighting Inc., 382 F.3d 1354, 1358 (Fed. Cir. 2004).

[3] Williamson v. Citrix Online LLC, 2013-1130, slip op. at 16 (Fed. Cir. Jun. 16, 2015) (en banc).

[4] Linear Tech. Corp. v. Impala Linear Corp. , 379 F.3d 1311, 1320-21 (Fed. Cir. 2004) (“We hold that because the term ‘circuit’ is used ... with a recitation of the respective circuit’s operation in sufficient detail to suggest structure [that the ‘circuits’] are not means-plus-function limitations.”); Inventio Ag v. Thyssenkrupp Elevator Am., 649 F. 3d 1350, 1359 (Fed. Cir. 2011) (“computing unit” not means-plus-function).

[5] Ex Parte Smith, 2012-7631 (PTAB Mar. 14, 2013); Ex Parte Lakkala, 2011-1526 (PTAB Mar. 13, 2013); Ex Parte Erol, 2011-1143 (PTAB Mar. 13, 2013);

[6] Alice Corp. Pty. Ltd. v. CLS Bank Intern., 134 S.Ct. 2347, 2355 (2014)

[7] Williamson, Slip Op. at 17 (“this passage [is] in a format consistent with traditional means-plus-function claim limitations. It replaces the term ‘means’ with the term ‘module’ and recites three functions performed by the ‘distributed learning control module.’”).

[8] Alice, 134 S.Ct. at 2355.

[9] Williamson, Slip Op. at 18.

[10] No. 2014-1048, -1061, -1062, -1063 (Fed. Cir. June 23, 2015).

[11] Id., Slip Op. at 10.