It is generally stated in that one must both show ownership of and give the public the invention in order to get the monopoly rights of a claim (e.g. 112 in US, Art. 83 EPO).

However, for purposes of validity, is it acceptable to make a broad genus claim (e.g. all filamentous lightbulbs), but only disclose a few embodiments (e.g. with a horse hair filament), even when that scope (at least on its face) would include other embodiments that were unknown and/or required undue experimentation to be operably realized at the time of disclosure (e.g. carbon nanotubes as filament)? See the diagram below.

enter image description here

Note that the question does not relate to instances where an additional limitation is added (e.g. a frosted lightbulb), but to the more fundamental case where the claimed limitation (e.g. a filament) is satisfied with an embodiment unknown at the time of disclosure. In other words, the typical "three legged chair covers a four legged chair" would not apply.

I often see such broad non-enabled/non-sufficiently disclosed claims. However, the spirit of the law does not seem to support this, and e.g. this wikipedia article agrees:

Patents may fail this test if they claim more than they teach--for example, a patent that claims all light bulbs but that only explains how to make a particular type of light bulb.

In a previous question, I asked generally about infringement of such broad claims, the answer of which was, paraphrasing, "claim construction is limited to the intent of the claim terms". The question I want to ask here is: are such broad claims invalid when an unenabled/non-writtenly-described embodiment is identified? This is in the spirit of, "claims should be taken at face value" in the context of validity.

  • I think the distinction between USPTO and EPO is very impotant here, for which are you asking?
    – user18033
    May 4, 2018 at 17:01
  • If both, would you mind making this two questions and linking them? I think each jurisdiction allows for a long answer on it's own and most people will know about only one in depth.
    – user18033
    May 4, 2018 at 17:02
  • @DonQuiKong Feel free to answer either and notate accordingly. I actually don't expect it to be much different, though probably stricter in EPO.
    – user132162
    May 5, 2018 at 0:39

2 Answers 2


Are such broad claims invalid when an unenabled/non-writtenly-described embodiment is identified?


The written description requirements in the US and EPO require that the invention must be enabled or sufficient described (respectively) across the whole of its scope.

The USPTO sets this out reasonably clearly at MPEP § 2164.02:

For a claimed genus, representative examples together with a statement applicable to the genus as a whole will ordinarily be sufficient if one skilled in the art (in view of level of skill, state of the art and the information in the specification) would expect the claimed genus could be used in that manner without undue experimentation.

The EPO is the same at the core (though perhaps different at the edges). This is neatly summarised by the EPO Guidelines at F-III, 1:

There are some instances where even a very broad field is sufficiently exemplified by a limited number of examples or even one example (see also F‑IV, 6.3). In these latter cases the application must contain, in addition to the examples, sufficient information to allow the person skilled in the art, using his common general knowledge, to perform the invention over the whole area claimed without undue burden and without needing inventive skill (see T 727/95). In this context, the "whole area claimed" is to be understood as substantially any embodiment falling within the ambit of a claim, even though a limited amount of trial and error may be permissible, e.g. in an unexplored field or when there are many technical difficulties (see T 226/85 and T 409/91).

As a consequence, if a broad claim includes an unenabled embodiment, then by definition this means that the broad claim is not enabled over its full scope. The claim therefore is prima facie invalid.

As an aside, a claim which is prima facie invalid in such a way may be construed so as to be valid by excluding non-enabled species. This occurs somewhat more in the US, where implicit limitations are more frequently used.

But what about claims that include a generic component?

If a claim includes a temperature sensor, then this is enabled if the skilled person could perform the invention (that is, could identity and use a temperature sensor). This is particularly the case if it doesn't matter how the temperature sensor functions, just that it functions.

Because of this, it doesn't really matter if a novel and non-obvious temperature sensor is later invented and used by an infringer with the other components of the claim: the claim only requires a temperature sensor (and is enabled for any kind of temperature sensor), and so a use of a new temperature sensor would equally infringe.

However, consider an invention that required a temperature sensor that measured temperatures in a range of 0K to 1000K. In this case, it does matter how the temperature sensor functions (in the sense that it must meet certain criteria). If we only described example temperature sensors that measured 100K to 200K, 400K to 600K, and 700K to 1000K, then the invention is not enabled across the whole scope. That is, the skilled person could not perform the invention across its whole scope, because they could not identify a suitable temperature sensor for certain parts.

This doesn't change over time: it is assessed at the application date. If it's enabled on the application date, then it's enabled.

  • Okay, this makes sense, but the result is that technically, a lot of patent claims are invalid over lack of broad, non-specific claim language. E.g. see the bottom of this answer that refers to a Patently-O article: patents.stackexchange.com/a/19487/18769
    – user132162
    May 8, 2018 at 19:32
  • Also, another point of this question (assuming your answer is correct) is that patents which were valid at the time of filing (due to no known non-enabled/writtenly-described embodiments within the claim scope) may become invalid at a later date if technological progress identifies new embodiments that fit under the literal claim language, but were not sufficiently disclosed.
    – user132162
    May 8, 2018 at 19:34
  • Sorry to go on further here, but I just want to verify this example: if there are many patent claims in an art area including the limitation of a "sensor", and one comes up with a new sensor that is inventive over the prior art, are all those claims invalid due to failing full-scope enablement, and/or does practicing the issued claims with the new "sensor" not infringe?
    – user132162
    May 12, 2018 at 18:45
  • 1
    @user132162 I have revised my answer, which should hopefully address your point.
    – Maca
    May 13, 2018 at 0:25
  • 1
    @user132162 That passage merely confirms the difference between the predictable arts (like mechanical things) and the unpredictable arts (like pharmaceuticals) in terms of the level of specificity required. In unpredictable arts, you can't easily know what members are inside a genus. So you can't claim a result when there's no way to predict what specifics you are actually claiming. This does not have any bearing on when enablement is assessed or anything like that.
    – Maca
    May 19, 2018 at 1:34

In the US case, I would also like to add that reverse doctrine of equivalents (DOE) may apply as implicit claim scope limitation. Reverse DOE essentially says literally infringing structure may not infringe if it does not satisfy all of the traditional function-way-result requirements. E.g., see the wikipedia article.

This would apply to the previous question cited in the OP, i.e. the magic wand infringing a "wooden handled instrument for cutting down trees", as it would perform the function in a substantially different way than an ax.

  • 1
    Don't know if biotech patents comment on this, but when I was looking at that subfield, it seemed clear that new uses for an existing technique are patentable.
    – DukeZhou
    May 8, 2018 at 19:52

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .