The Federal Circuit’s en banc decision in Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015), expanded the potential application of 35 U.S.C. §112, ¶6, making it more likely that functional claim language will be construed as a means-plus-function limitation even in the absence of the word “means.” Patent claims that recite functions in connection with nonce words like “module,” “mechanism,” “element,” “device,” or even “processor” are now more likely to be deemed means-plus-function limitations. Whether a claim term is or is not subject to §112, ¶6 may be dispositive in some patent cases. For example, the specification must disclose a structure or algorithm for performing the means-plus-function limitation, and if no such structure is disclosed, the claim will be held invalid as indefinite. Recent cases applying Williamson have reached different results, with some decisions finding claims subject to §112, ¶6 and invalid for lack of structure and other decisions finding software claim terms to recite structural limitations not subject to §112, ¶6. While most of the decisions to date have been in the computer-related arts, interesting parallels exist in the life sciences and pharmaceutical fields. Below, we discuss recent decisions applying Williamson and provide practical insights and strategies for patent owners and accused infringers to consider when addressing the expanded application of §112, ¶6.
Federal Circuit Changes Standard for Application Of §112, ¶6 in Williamson
Patent law generally does not permit purely functional claiming because this would allow a patent owner to obtain greater coverage without sufficiently describing the invention. Title 35 U.S.C. §112, ¶6 provides an exception for “means-plus-function” claims. A claim limitation can be “expressed as a means or step for performing a specified function without the recital of structure … in support thereof, and such claim shall be construed to cover the corresponding structure … described in the specification and equivalents thereof.” 35 U.S.C. §112, ¶6 (now renumbered §112(f)). Whether the word “means” is present in the claim is important. Use of the word “means” triggers a rebuttable presumption that §112, ¶6 applies; absence of the word “means” triggers the opposite rebuttable presumption. For more than a decade, the Federal Circuit characterized the presumption against application of §112, ¶6 when the word “means” is absent as a “strong one that is not readily overcome.” The court said it has “‘seldom’ held that a limitation without recitation of ‘means’ is a means-plus-function limitation” and refused to apply §112, ¶6 unless the claim “essentially is devoid of anything that can be construed as structure.” E.g., Lighting World, Inc. v. Birchwood Lighting, Inc., 382 F.3d 1354, 1358 (Fed. Cir. 2004); Apple, Inc. v. Motorola, Inc., 757 F.3d 1286, 1297 (Fed. Cir. 2014).
Williamson overruled this heightened bar to overcoming the presumption against applying §112, ¶6 to functional language lacking the word “means,” which generally made it easier to show that §112, ¶6 applies. Williamson, 792 F.3d at 1349. The Federal Circuit changed the standard because it found the “strong presumption” against applying §112, ¶6 to functional limitations without the word “means” to be unwarranted and to have resulted in a proliferation of functional claiming that does not follow the statutory requirements of §112, ¶6. To determine whether §112, ¶6 applies, Williamson announced the “standard is whether the words of the claim are understood by person of ordinary skill in the art to have a sufficiently definite meaning as the name for structure.” The presence or absence of the word “means” still creates rebuttable presumptions for (if “means” is present) and against (if “means” is absent) application of §112, ¶6, but now the §112, ¶6 presumptions are in balance and there is no thumb on the scale disfavoring application of §112, ¶6. The presumption against §112, ¶6 can be overcome by showing merely that the claim “fails to ‘recite sufficiently definite structure’ or else recites ‘function without reciting sufficient structure for performing that function.’”
Opportunities And Challenges Presented by Williamson
The revised standard for applying §112, ¶6 has resulted in more claims without the word “means” being construed as means-plus-function claims, and it has caused some claims to be invalidated as indefinite at the claim construction phase for failing to disclose a corresponding structure. For example, the patent at issue in Williamson was directed to a virtual classroom system that connected presenters with geographically remote audiences. As part of the virtual classroom, the claims recited “a distributed learning control module for receiving communications transmitted between the presenter and the audience member computer  and for relaying the communications to [a] receiving computer system and for coordinating the operation of the streaming data module [that streams selected data to the presenter's and audience member's computers].” The Federal Circuit upheld application of §112, ¶6 to both of the “module” claim terms, even though they did not use the word “means,” because the term “module” did not indicate any particular structure and was determined to be “a well-known nonce word that can operate as a substitute for ‘means.’” Expanding on the revised test, the en bancWilliamson court also observed that generic terms such as “mechanism,” “element,” “device,” and other nonce words that are just verbal constructs may be synonymous with the word “means” because they “typically do not connote sufficiently definite structure.” Because §112, ¶6 applied to the “module” terms, the court looked to the specification to construe the terms as covering the structure linked or associated with the functions performed by the “modules.” For computer-implemented means-plus-function terms, the specification must disclose an algorithm or series of steps for performing the claimed function. But no such structure or algorithm was disclosed for all of the functions of the “module” terms. The Federal Circuit also rejected as irrelevant the testimony of the patent owner’s expert that a person of ordinary skill “would know exactly how to program” the recited “modules,” explaining that expert testimony cannot create structure where none is disclosed. Because §112, ¶6 requires disclosure of a corresponding structure and because no corresponding structure was disclosed for the “module” terms, the Federal Circuit agreed the “module” terms rendered the claims invalid for indefiniteness.
The Federal Circuit and other courts have applied Williamson to construe under §112, ¶6 a growing class of claim terms lacking the word “means” and to hold some of those claims invalid as indefinite for failing to disclose sufficient corresponding structure. These cases are significant because they have invalidated patents relatively early in litigation at the claim construction stage. In Media Rights Technologies, Inc. v. Capital One Financial Corp., 800 F.3d 1366, 1372 (Fed. Cir. 2015), the patent was directed to a system for preventing unauthorized recording of content using a “compliance mechanism” to determine whether the content was authorized before publishing it. The recited “compliance mechanism” term, though lacking the word “means,” was determined to be subject to §112 ¶6 under the Williamson test, because the term does not refer to a specific structure and all of the specification’s examples were merely illustrative and were described only in functional terms. The Federal Circuit also found the patent indefinite because the specification failed to disclose structures corresponding to each of the functions of the “compliance mechanism” recited in the claims. The Federal Circuit reached similar conclusions in Advanced Ground Information v. Life360, Inc., 830 F.3d 1341, 1347-48 (Fed. Cir. 2016), which involved patents directed to cellular communication systems that monitor and display the location of multiple users on a map together with other information. The claim term at issue was “symbol generator” and this term did not include the word “means,” but the court applied §112, ¶6 because a “symbol generator” that generates symbols representing each user in the network on the map is analogous to a “means for generating symbols” and merely recites the function performed without identifying a specific structure for performing it. The Federal Circuit reached this conclusion even though the terms “symbol” and “generator” individually were well-known terms, and even though an expert testified a person of skill would know how to select and use the recited “symbol generator” from a known class of software. Because §112, ¶6 applied, the court looked to the specification to identify an algorithm to supply structure to perform the “symbol generator’s” functions, but found no such algorithm and thus found the claim indefinite and invalid.
The exact claim language at issue is critical to the §112, ¶6 analysis. One court found the word “module” in a computer-implemented claim to a transaction system not subject to §112, ¶6 under Williamson because the claim recited in detail how the “module” performed the software functions. St. Isidore Research, LLC v. Comerica Inc., No. 15-cv-1390, 2016 WL 4988246, at 13 (E.D. Tex. Sept. 19, 2016). Other courts have applied Williamson to terms that appear to recite pure functionality or pure software, but have found the terms to be structural, and therefore not subject to §112, ¶6. For example, “code segment” was the term at issue in Collaborative Agreements, LLC v. Adobe Systems, Inc., No. 15-cv-3853, 2015 WL 7753293, 4-5 (N.D. Cal. Dec. 2, 2015). “Code segment” was indisputably software, but the code segment’s operations were recited in detail in the claims (e.g., receiving documents and identity validation information from a first party that confirms the first party has agreed to the documents, posting the agreed-to documents to an account on a server accessible to a second party). The court first found that “code segment” identified a type of structure, i.e., program instructions, and then found the claims themselves supplied sufficient detail about how those program instructions operated, which was enough to maintain the presumption that “code segment” was not subject to §112, ¶6, even though “code segment” is a functional term. The court analogized the term “code segment” to the term “circuit,” which can also connote sufficient structure when combined with a description or drawing showing how the circuit operates. The accused infringers also argued an algorithm explaining exactly how, e.g., the code segment did the “receiving” with specific instructions was required to avoid application of §112, ¶6. But the court rejected this argument as going beyond the threshold trigger for determining whether §112, ¶6 applies and asking whether the term was enabled or whether a sufficient structure was disclosed for a term already deemed subject to §112, ¶6.
Some courts have also used similar analysis to find the term “processor for … [performing a function]” to be structural and not subject to §112, ¶6. See, e.g., Advanced Mktg. Sys., LLC v. CVS Pharm., Inc., Case No. 15-cv-134, 2016 WL 1741396, at 19-20 (E.D. Tex. May 3, 2016) (“data processor for [performing a function recited in detail]” not subject to §112, ¶6 because “ processor” is structural and the claims described how the processor performs the functions); Smartflash LLC v. Apple Inc., 77 F. Supp. 3d 535, 543 (E.D. Tex. 2014), reconsideration denied, 2015 WL 4208754, at 3 (E.D. Tex. July 7, 2015) (“processor” and “code” are structural terms whose operation was sufficiently described even in light of Williamson). But other courts considering the term “processor” have reached opposite results. See, e.g., St. Isidore, 2016 WL 4988246, at 14-15 (“processor configured to [verify with text messages whether a user is authorized to access an account]” not structural and thus subject to §112, ¶6 because the term was only defined by its function); Personal Audio, LLC v. Apple, Inc., No. 09-cv-111, 2011 WL 11757163, at 21-22 (E.D. Tex. Jan. 30, 2011) (“processor” subject to §112, ¶6 under the more difficult pre-Williamson standard where two patents were essentially identical, except one used the word “processor” in place of “means”).
Potential Application of Williamson to Life Sciences And Pharmaceutical Patents
Williamson has primarily impacted computer-related cases and has not yet filtered into life sciences and pharmaceutical cases. Functional claiming arguments in life sciences and pharmaceutical cases often rely on challenges to written description and/or enablement. See, e.g., Regents of the Univ. of Calif. v. Eli Lilly & Co., 119 F.3d 1559, 1568 (Fed. Cir. 1997) (rejecting claim “because it does not distinguish the claimed genus from others, except by function. It does not specifically define any of the genes that fall within its definition … . It is only a definition of a useful result rather than a definition of what achieves that result.”). However, as compared with §112, ¶6, the enablement and written description requirements are less burdensome and thus they are less likely to result in an invalidity finding when functional claim language is used. Further, because life sciences and pharmaceutical patents are more likely to use claim terms that identify a particular structure or type of structure, it may be difficult to win an argument for applying §112, ¶6 unless the word “means” is used, which may partially explain why these arguments have not been raised.
In light of Williamson, however, there are some interesting parallels between computer-related and life sciences/pharmaceutical terms alleged to be purely functional. For example, a claim reciting an “antibody” that binds to certain receptors and/or blocks certain bindings can be analogized to a “processor” for performing a given function. The term “antibody” identifies a type of structure, but if the claim or specification fails to describe how the antibody functions in sufficient detail, it might be possible to ask for a §112, ¶6 construction. Such a construction would limit the term to the corresponding structure in the specification, which might significantly strengthen an accused infringer’s argument for non-infringement, or even invalidate the patent if there is no corresponding structure in the specification for a function the “antibody” performs.
Practical Strategies for Accused Infringers and Patent Owners
Accused infringers should look to Williamson as presenting more opportunities to substantially narrow or potentially invalidate patent claims as indefinite. The strong presumption against §112, ¶6 no longer applies to claims lacking the word “means.” And an accused infringer can establish by a preponderance of the evidence that claim terms are subject to §112, ¶6.
Claim terms subject to §112, ¶6 will be construed to cover the corresponding structure in the specification, which generally results in a narrower construction that may not be infringed, and if no such structure is disclosed, the claim will be indefinite and invalid. This indefiniteness test is easier to meet than showing with clear and convincing evidence that a structural claim term fails to inform a person of skill of the claim’s scope with “reasonable certainty.”
Claim terms such as “module,” “mechanism,” “element,” “device,” or even “processor” should be closely examined for potential means-plus-function arguments. Likewise, other nonce words that do not denote a specific structure or type of structure should be examined closely, especially words coined by the patent owner for purposes of the patent.
Patent owners who have not mounted sufficient defenses to a means-plus-function attack or who have been unable to avoid means-plus-function constructions for some terms have had their patents invalidated. See, e.g., Zeroclick, LLC v. Apple, Inc., No. 15-cv-4417, 2016 WL 5477115, at 4, 6 (N.D. Cal. Aug. 16, 2016) (invalidating claims as indefinite for failing to disclose corresponding structure under §112, ¶6 where the patent owner did not attempt to identify any related structure recited in the claims and failed to explain why the term “code” referred to a specific structure).
Patent owners can no longer merely rely on the absence of the word “means” to defeat an accused infringer’s argument that a functional claim term is not subject to §112, ¶6. They must be prepared to articulate why a claim term is structural based on the claim language and intrinsic evidence.
Patent owners should also consider using expert testimony where appropriate to show a challenged claim term refers to a known structure.
Patent owners should not concede a claim term is functional, especially where it will be difficult to show that a corresponding structure (or algorithm for computer-implemented inventions) is disclosed in the specification for all of the functions performed in the claim. Lack of structure for just one recited function can result in an invalidity finding.
The increased scrutiny Williamson has focused on functional claim language has resulted in more claim terms being subjected to §112, ¶6 analysis and has invalidated patents that may not have been invalidated under the standard applicable before Williamson.
Joshua D. Curry is a Partner in the Atlanta office of Dentons US LLP. His practice focuses on patent and IP litigation and pre-litigation counseling involving complex technologies. He can be reached at firstname.lastname@example.org. Kate E. Hart is an Associate in the Kansas City office of Dentons US LLP. Her practice focuses on IP litigation and prosecution. She can be reached at email@example.com.
The views expressed in the article are those of the authors and not necessarily the views of their clients or other attorneys in their firm.