Written description decisions are unusually fact intensive. The court must answer (1) what exactly was invented? and (2) what exactly was claimed? Then the court must dissect the differences between the invention and the claims.
A patent has an adequate written description if it “reasonably conveys to those skilled in the art that the inventor had possession of the claimed subject matter as of the filing date.”
In Rivera v. ITC, the patent concerned a coffee brewer that could accept both “cup” and “pod” containers. A “cup” is a small, self-contained plastic cartridge. A “pod” is a small disc-shaped filter containing coffee grounds. Cups and pods hold coffee grounds differently, so it’s a bit of a trick to develop a brewer that can use both.
But Rivera took their claim on a cup-or-pod brewer and asserted that a loose-grounds-inside-a-cup apparatus infringed. This was too far for the ITC and ultimately the Federal Circuit panel:
[T]he question is whether a pod adaptor assembly intended to allow compatibility between distinct brewing systems, also supports an undisclosed configuration that eliminates a fundamental component of one of those systems (i.e., the “pod”) through integration. It does not.
I’ve long supported meaningful enforcement of the written description requirement: Inventors should get patents on what they actually invented. The written description requirement helps enforce that basic (and fair) requirement.
Back in November 2016, the Federal Circuit narrowed the scope of patents eligible for Covered Business Method review. They held that a patent is not eligible for CBM review just because it could involve a financial transaction. Instead, the patent must have a claim that actually contains a financial activity element.
The Federal Circuit’s 2-1 decision in Secure Axcess v. PNC Bank reaffirms that narrowing, and adds that the litigation history of the patent (here targeting financial institutions) is also not relevant:
[A] patent owner’s choice of litigation targets could be influenced by a number of considerations, such as the volume of a particular target’s perceived infringement; the financial condition of the target; which targets are most likely to be willing to settle rather than bear the cost of litigating; available and friendly venues; and so on.
But Judge Lourie dissented, arguing that the patent specification and litigation history clearly described a patent on a “financial product or service,” regardless of whether the claims specifically include a financial transaction.
Will we have a bright-line rule, or an “all the circumstances” test? So far the bright-line rule has the upper hand.
A complicated decision by the Federal Circuit gives us all an opportunity to review the MedImmune standard. Prior to MedImmune a licensed party could not bring an invalidity suit because it need not fear infringement (i.e., no case or controversy). MedImmune loosened that standard and essentially set forth a “look at all the circumstances” test.
So what if a party kinda asserts patent claims, but it’s really not clear, and then later gives a covenant not to sue. Somewhat predictably the “all the circumstances” test leads to a 2-1 split in ArcelorMittal v. AK Steel. The panel ends up disagreeing on whether the circumstances of a “covenant not to sue” moot the underlying controversy.
Aaaaaand… that’s really all you need to know unless you want the complicated details of this case.
In Phigenix, Inc. v. Immunogen, Inc., a Federal Circuit panel concluded that a company can have standing to initiate an IPR against a patent, but not have standing to appeal the results of that IPR to the Federal Circuit. This standing gotcha arises because the Article III “case or controversy” requirement requires that a party:
must have (1) suffered an injury in fact, (2) that is fairly traceable to the challenged conduct of the [appellee], (3) that is likely to be redressed by a favorable judicial decision.
In this case Phigenix was a third-party biotechnology company in the same space as the patent owner, and initiated an IPR that the PTO ultimately closed after concluding that the claims were not obvious. Phigenix then of course sought to appeal that decision to the Federal Circuit under the relevant statute that says, “A party to an inter partes review . . . who is dissatisfied with the final written decision of the [PTAB] . . . may appeal the [PTAB]’s decision only to the . . . Federal Circuit.”
Not so fast. The Federal Circuit held that Phigenix could appeal the result, but that didn’t mean it had standing to actually get a decision:
Phigenix does not contend that it faces risk of infringing the ’856 patent, that it is an actual or prospective licensee of the patent, or that it otherwise plans to take any action that would implicate the patent.
Phigenix tried some other arguments surrounding standing, but the Federal Circuit methodically shut these down as well based on relatively solid Supreme Court jurisprudence. It’s hard to say this case is wrong given the case law, but it is no doubt frustrating to see patent decisions dodging the merits of patent validity. These decisions have major public consequences even if they are (very) technically advisory with respect to the advocating party.
The DMCA safe harbor (17 U.S.C. § 512) immunizes websites and other internet service providers from copyright infringement caused by their users if the providers follow certain procedures. Generally this means providers have to remove copyrighted works as soon as they become aware of them.
But what if you have a bunch of volunteer moderators / curators determining which user submissions to post? Are you responsible for copyright infringement if the volunteers approve a copyrighted user submission? It seems likely, if those volunteers qualify as “agents” of your company.
In Mavrix Photographs v. LiveJournal, Inc. (April 2017), a Ninth Circuit panel concluded that volunteer moderators on a website could indeed be “agents” of the website, although the agency determination is a question of fact that cannot be resolved on summary judgment.
Importantly, the panel ruled that curating the posts (i.e. only posting about 1/3 of the submissions) probably didn’t qualify LiveJournal for the safe harbor of posting “at the direction of the user”. Previous decisions have allowed websites to take advantage of the safe harbor so long as they were only making “accessibility enhancing” changes to posts such as reformatting or screening for offensive material. But curation may be too much control:
The question for the fact finder is whether the moderators’ acts were merely accessibility-enhancing activities or whether instead their extensive, manual, and substantive activities went beyond the automatic and limited manual activities we have approved as accessibility-enhancing.
This reading appears to be well within the statutory language, but it threatens to put additional burdens on websites that curate (and begs the question of what exactly is curation / moderation). If you curate, how can you also ensure the content is not copyrighted by someone else? It’s not obvious.
Update: A group of internet companies including Google and Facebook have urged en banc review. They argue that this decision incentivizes companies to do no pre-screening of user submitted material, and that this is bad policy and contrary to the intent of the DMCA.
This all boils down to whether some forms of pre-screening are extensive enough to consider it reasonable that the company should police copyright infringement too. And to be fair the panel’s decision didn’t say one way or the other: it just said there are enough fact issues that you can’t decide on summary judgment.
From the New York Times:
Hackers exploiting data stolen from the United States government conducted extensive cyberattacks on Friday that hit dozens of countries, severely disrupting Britain’s public health system and wreaking havoc on tens of thousands of computers elsewhere, including Russia’s ministry for internal security.
There are really only two things that need to be said about this, both said well by others:
The point is there’s no such thing as a security backdoor that “only I can use.” If you want systems to truly be secure, they must truly be secure.
Sonix owns a patent on using a “graphical indicator” on a surface such as a children’s book to subtly encode information about the surface. It can be used with an optical reader to allow the reader to play, for example, a pig sound when placed over the image of a pig.
The claims say that the graphical indicator has to be “visually negligible.” When Sonix sued someone on that patent, they (eventually) said that term was indefinite.
The district court agreed, and as seems to be common in these indefinite cases, the Federal Circuit reversed:
a skilled artisan would understand, with reasonable certainty, what it means for an indicator in the claimed invention to be “visually negligible.”
The decision cites examples of visually negligible indicators called out in the patent specification. But the decision also states that the parties’ own behavior during the litigation supports the result.
I’m not a fan of citing litigant behavior in support of legal conclusions. Litigants make many complex decisions about which arguments to pursue and which to drop, and it’s impossible to reconstruct that calculus on a cold record. Courts should stick to the law and focus on providing predictable guidance on that basis. But it does seem common for courts to cite such behavior so you have to be careful.
Ameranth owns a few patents related to the automatic generation of menus, like in a restaurant setting where subsequent menu choices change depending on initial selections. These patents got involved in a Covered Business Method review by the PTO, which found that some of the claims were unpatentable under section 101.
On appeal to the Federal Circuit, the panel decided that in fact all the claims were unpatentable under 101.
Some claims already ineligible. The PTO had decided that some of the claims were directed to the abstract idea of generating a second menu from a first menu, and sending the second menu to another location. According to the Federal Circuit panel, the claims:
Thus the claims were generic enough to qualify as an abstract idea.
Any other claimed steps (input/output, network communications, etc.) were “insignificant post-solution activity” of a type already known in the case law.
Even more claims ineligible. Most significantly, the PTO had found a few claims eligible because they required linking a menu to a specific customer at a specific table. The PTO had decided this functionality was novel (at least for mobile devices) in 2001. But the Federal Circuit reversed, citing in part a concession made at oral argument that restaurants have always been able to keep track of which customer at which table ordered which meal: “Merely appending this preexisting practice to those independent claims does not make them patentable.”
My views. On the whole, I agree with the outcome here: these claims are generic, don’t solve a truly technical problem, and simply involve automation of well-known activities. But as a side note, I dislike some formulations of this argument.
Petitioners argued that the table-linking claims were ineligible because they were a “classic example of manual tasks that cannot be rendered patent eligible merely by performing them with a computer.” There are many, many manual tasks performed by humans that would be amazing if computers could do them. We go much too far if we make this kind of argument without the appropriate qualifications.
Covered Business Method (CBM) Patent Review is a special PTO procedure designed to allow reexamination of a certain class of patents popularly criticized for their overbreadth and lack of “true innovation.” These are the so-called “business method patents.”
By statute a patent is eligible for CBM review if it claims a method for “data processing . . . used in the practice, administration, or management of a financial product or service” but specifically excepting patents for “technological inventions.” This is sort of vague, and that leads us to the Unwired Planet v. Google decision.
Unwired Planet sued Google on a patent for restricting an app’s access to location information. Google asked the PTO to reexamine the patent in a CBM proceeding because the patent says the app might be associated with sales services such as a restaurant or store. The PTO decided that the patent was in fact “incidental or complementary to the financial activity” because location services could involve an eventual sale of services.
This analysis didn’t work for the Federal Circuit panel.
The main problem was the PTO never adopted this “incidental or complementary” language in a formal regulation. That language isn’t in the statue, and applying it now against Unwired Planet was improper.
But perhaps more importantly, the panel emphasized:
[I]t cannot be the case that a patent covering a method and corresponding apparatuses becomes a CBM patent because its practice could involve a potential sale of a good or service. All patents, at some level, relate to potential sale of a good or service.
Thus, “[i]t is not enough that a sale has occurred or may occur, or even that the specification speculates such a potential sale might occur.”
Net result: the PTO should not have accepted the patent for CBM review. Reversed and remanded.