Narrowly Claimed, Broadly Disclosed: When Patent Claims Mean Exactly What They Say

In a short but instructive opinion , the U.S. Court of Appeals for the Federal Circuit delivered a reminder about the relationship between claim language, written descriptions and the limits of what a patent’s claims can cover. Dynapass owns U.S. Patent No. 6,993,658 , covering methods for authenticating users on a secure computer network using a two-component credential: a passcode that the user knows, and a token that the user receives via cell phone. The key idea is that these two pieces are combined into a single password , which the user then submits to receive access. This security protocol is of a form frequently referred to as “multi-factor authentication” (MFA). Dynapass alleged that the MFA feature in the defendant’s mobile banking app infringed claims 1-7 of the ’658 patent. The dispute ultimately turned not on whether the mobile app used tokens and passcodes, but on the meaning of “receiving the password.” Claim 1 of the ’658 patent recites a sequential method: (1) generating a new password based on the token and the passcode being combined; (2) setting that new password on the user’s account; and (3) receiving the password from the user. The district court construed the final step of “receiving the password” as requiring receipt of the combined token and passcode. Thus, the system had to receive the token and passcode already merged into a single credential; receiving them separately would not qualify. Dynapass appealed, arguing that the ’658 patent’s written description discloses an alternative embodiment in which “the passcode 154 and the token 156 are submitted separately.” Dynapass argued that the district court had improperly limited the claim scope to the recited “combined” embodiment, and ignored this “submitted separately” alternative, thereby excluding an embodiment that the patent explicitly described. The Federal Circuit was not persuaded, affirming the district court’s construction based on the plain language of the claims. The claim language itself, the court explained, forecloses Dynapass’s reading because the claim’s sequential structure is unambiguous: first generate a password from the token and passcode, then receive the password from the user. The phrase “the password” is a clear reference back to the newly generated, combined credential, rather than its constituent parts. Receiving a passcode and a token in isolation simply is not equivalent to receiving the password those components would form when combined. The Court also found that the written description reinforced this reading, as it never describes the alternative embodiment’s separate submission as “receiving the password.” Instead, the specification distinguishes between “password” and “passcode and token data.” The Court cited this pattern as particularly meaningful, noting “that when a patent repeatedly and consistently characterizes a claim term in a particular way, it is proper to construe the claim term according to that characterization.” The court also cited TIP Systems, LLC v. Phillips & Brooks/Gladwin to counter the argument that the claims had been improperly limited to a disclosed embodiment, invoking the principle that the existence of an alternative embodiment in the specification does not automatically broaden the claim to cover it: the patentee chose the claim language, and that language controls. Thus, a patentee cannot draft narrow claims, and then argue for a broader interpretation based on the description to cover accused products that the claims as written would not otherwise cover. While not at issue in Dynapass , similar situations involving claims of infringement under the doctrine of equivalents may implicate, and be limited by, the disclosure-dedication rule. The disclosure-dedication rule provides that when a patent’s written description discloses subject matter that the patentee does not claim, that unclaimed subject matter is dedicated to the public. The patentee cannot later recapture it through the doctrine of equivalents, allowing a patent to cover products or processes that don’t literally infringe a claim, but are only insubstantially different from what is claimed. The disclosure-dedication rule exists for good reason: the patent laws require inventors to disclose their inventions in exchange for a monopoly extending only as far as the claims. If a patentee describes a broad invention, but then claims only a narrow embodiment of it, the public is entitled to use the unclaimed portions. Allowing the patentee to recapture those unclaimed portions through equivalents would undermine that reliance, and effectively reward inaccurate or vague claim drafting. Although the ’658 patent’s written description explicitly discloses the alternative embodiment where “the passcode 154 and the token 156 are submitted separately,” the claims were written to require a unified password. The alternative embodiment was described in the specification, but was never claimed. If Dynapass had attempted to argue that the defendant’s separate-component submission was nevertheless equivalent to “receiving the password,” the disclosure-dedication rule would have barred that argument. The separately-submitted-components embodiment was affirmatively disclosed and not claimed, and so belongs to the public. This opinion illustrates an important concept for patent drafting: Describing alternative embodiments in the specification is potentially advantageous since it can support claim breadth and demonstrate enablement. However, every unclaimed embodiment is a potential gift to the public under the disclosure-dedication rule. Patentees who draft broad descriptions, but then pursue narrow claims may find that both literal infringement and the doctrine of equivalents are unavailable to cover the very alternatives they chose to describe. Drafting broader claims, or, where broad claiming isn’t possible, being cautious about which alternative embodiments to disclose, can help avoid inadvertently surrendering them.