Risks of "kitchen sink" disclosures?

Question

It is common practice when writing specifications to disclose a broad range of possible embodiments "just to be safe", and without much thought to how such embodiments might be realized in practice. However, is there a downside to doing this? Without thinking through and verifying that all embodiments are enabled (e.g. all combinations and subcombinations of features disclosed), it seems possible that generic claims encompassing the disclosure would be invalid over enablement. E.g. see here:

The claim was broad enough to cover both mechanical and electronic sensors. The patent's specification, on the other hand, described mechanical sensors in detail, however, the electronic sensor was described in broad terms and illustrated in one conceptual figure. The Court of Appeals ruled that the claim was invalid on the grounds that the full scope of the claims covered electronic sensors as well as mechanical sensors, but the specification did not teach a person skilled in the art how to make and use the invention with an electronic sensor (i.e. the claim was not enabled).

Note: This also could bring up the issue of operability/"utility" of embodiments within the claim scope (e.g. some combinations of disclosed features may result in a nonsense embodiment).

---

Edit: Here's an example to illustrate the question. Assume you are disclosing a novel spring mechanism for a stapler. In the description, you write:

In some embodiments, the stapler may comprise an anvil to provide a counter-force with respect to the arm during a stapling operation. In other embodiments, the stapler may not comprise an anvil to provide a counter-force with respect to the arm during a stapling operation, e.g. the counter-force may be provided by an external object such as a cork board. In some embodiments, the stapler may be used for office tasks such as coupling paper. In some embodiments, the stapler may be used in surgical settings, e.g. for coupling skin flaps.

...

1. A stapler with [the novel spring mechanism].

Therefore, based on the explicit description, the claim scope presumably comprises at least the following four embodiments:

1. With anvil for office tasks
2. Without anvil for office tasks
3. With anvil for surgical settings
4. Without anvil for surgical settings

But how would 4 be enabled? It doesn't seem enabled to press a stapler without a bottom-part/anvil into flesh until a staple comes out. That would not achieve the desired purpose, and let's assume it does not for the sake of argument here. Therefore, is it implicit that the claim scope only includes embodiments which would be "reasonable" to a PHOSITA, thus excluding non-enabled embodiment 4 from the claim scope?

If it were the case that embodiment 4 was not considered part of the claim scope, what if someone sold the no-anvil embodiment with a separate "anvil plate" that was to be manually held in place to achieve the desired result in surgical settings, though (again assuming for the sake of argument) not in a way that would infringe embodiment 3 under doctrine of equivalents (same way, function, and result). It seems like the patentee would point to the broad claim and the spec to say that the separate anvil plate embodiment clearly infringed what would be embodiment 4, even though as described above, embodiment 4 does not seem to be enabled on its own.

Answer 1

Practice differs across jurisdictions worldwide. The idea is the same behind all of them, but usually the tools to analyze enablement are different.

I think patent attorneys don't like broad claims because they know it is tough to support them and the danger of having an invalid claim granted lures around. The applicant, unfortunately, is in most cases very happy with that and very unhappy with a narrow claim - until (and if) he sees his patent invalidated.

So yes, it is quite common to see broad claims drafted, but in my experience they rarely make it through examination. It all boils down to the person who drafted the application, whose job is to create enough and robust "falling back" positions in order to amend the claim(s) to a narrower version. There are different requirements depending on the jurisdiction, in US I think it is more liberal, while in Europe the safest road is to literally have the falling back positions drafted at least inside the description (to avoid the notorious Article 123(2) EPC objection). If it is too much, at least make sure that specific combinations are clearly suggested and disclosed in the text.

What really is enabling and what is not, usually is subjective. If it was e.g. a technique about boiling pasta and the disclosure was about spaghetti, penne and farfale, then you would expect that a claim to all pasta would be considered enabling. But a claim to pasta, rice or noodles, in view of the above examples, could be considered non-enabling, because rice is so much different compared to pasta and the skilled person might not be expected to be able to implement the new technique to rice in the absence of an example. This is exactly what a skilled patent attorney should be able to judge, at least by way of probability and in view of recent case law (you need to be updated, legal tests to define that change sometimes).

If you get an objection based on enablement during examination that's fine, if you are prepared - amend and go on. If you have a too broad claim granted, it depends on the flexibility of the procedure under which this claim is challenged. In Europe it is fairly easy to amend during opposition proceedings and keep your claim alive in a narrower form (but you need a nice and robust drafting for that as explained). In US it is not clear to me how easy it is to amend during IPR challenges, I think the situation is currently in a dynamic situation.

As a final remark, what is called "enablement" in US is "sufficiency of disclosure" in EPO, defined by Article 83 EPC.

Answer 2

Without going in detail, I want to point out some articles [1-3] below that address the risks of "kitchen sink" disclosures at least in the EPO, in the form of (a) intermediate generalization and (b) individualization.

[1]: International standards against intermediate generalization and individualization.

[2]: Individualization at the EPO.

[3]: Though in the US, individualization is “common, and often permissible”.

The overall point is that thoughtlessly including a long list of possible realizations for each feature without disclosing specific combinations of them could limit your ability to amend claims. Another way of putting this is, if you disclose everything (e.g. the entire contents of the internet), you've really disclosed nothing in particular. Therefore, there is a balance between disclosing enough to guarantee a potential infringer literally infringes on the claims (as opposed to weaker DoE infringement), and disclosing too much such that you can't amend claims to follow market development. Even worse, there is at least a US precedent of "disclosure-dedication", in which disclosed but unclaimed embodiments (i.e. those that technically fall outside of the claim scope), are not covered under DoE.

Therefore, best practices are to (i) as chempatent1981 notes, disclose the exact combinations of features that you may want to claim in the future (i.e. don't expect that the spec is a "bucket of legos" from which you can arbitrarily pick and combine elements, especially in the EPO), and (ii) don't get too crazy with long lists of possible embodiments, especially if they start to fall outside the scope of what you want to claim.

---

Also, in response to the stapler example given in the original question, an article from Patently-O makes an interesting comment that indeed, taken to the extreme, such general lack of specificity in claims could mean nearly all claims are invalid for 112/sufficiency of disclosure reasons:

In a powerful rhetorical approach, Judge Moore chose to refer to the broad claims as “spikeless claims.” As mentioned, those claims do not include a “spikeless” limitation. Rather, they simply omit a “spike” element. As it turns out more than 99.9% all patent claims issued in 2008 are silent about “spikes,” and under the traditional interpretation of the “comprising” transition – all those claims would literally cover embodiments without spikes. The holding here cannot be that all those claims are invalid. I believe that the holding here is largely a result of the fact that the accused device was in-fact spikeless.

In practice, it will likely not be ruled that "all claims are invalid", but rather that most claims will have a narrower scope than what they technically claim, narrowed both by the disclosure and PHOSITA knowledge.