Puzzles of Disclosure and Support

Paid Search Engine Tools, LLC v Google Canada Corporation 2021 FC 1435 McDonald J

2,415,167 / Paid Search Engine Bid Management

In my earlier post on this decision, I noted that McDonald J held that in light of the parallelism between 28.1(1), the claim date provision, and 28.2(1), the novelty provision, priority can only be claimed if the priority application discloses the same invention; it is not enough that the claimed subject matter is reasonably inferable from priority document [221]. Specifically, 28.1(1) says priority may only be claimed to a prior application “disclosing the subject-matter defined by the claim.” But as a comment on that post noted, 38.2 prohibits amendments that “add matter that cannot reasonably be inferred from the specification. . . on its filing date”; that means, conversely, that an amendment is permitted if the added matter can be reasonably inferred from the original application. Since the specification is part of the claim, it seems that 38.2 means a new claim is permitted so long as the subject-matter can be reasonably inferred from the original specification. On its face, there seems to be a gap between “reasonably inferred” and “disclosed” with the latter being more stringent than the former: see Western Oilfield 2021 FCA 24 on the difference between the Canadian and UK standards for added matter, discussed here. How can these differing standards be reconciled?

There is a straightforward answer, which is that 38.2 goes to amendments, and 28.1 goes to priority. So it is consistent to have a claim which is permitted (under the less stringent 38.2 standard), but which is not entitled to priority (under the more stringent 28.1 standard).

But problems lie not far below the surface. Consider a scenario in which the description in the application as filed is identical to that in the priority document. A claim is amended during prosecution in response to an objection by the examiner. The amended claim is reasonably inferable from the application as filed, so is permitted by 38.2, but it is not disclosed by the application as filed. The claim date for that claim must be either the filing date or the priority date: there are no other options. In particular, it is not an option under the Act or Rules for the claim date for that claim to be the date on which it was amended. (Compare the old Rule 53 under the old Act.) Section 28.1 provides that the claim date is the filing date unless priority is claimed. If the applicant does not claim priority, then the claim date is the filing date. The claim must also be supported by the description, per Rule 60. It must be implicit that if the amended claim can reasonably be inferred from the description, it is also supported by the description. Otherwise, the “reasonably inferred” standard for permitted amendment would be pointless, as any amended claim that was only reasonably inferred from the description would be invalid for lack of support. This is supported by Hughes J’s passing statement in Bartley 2011 FC 873 [15] equating “supported” with “reasonably inferred.” Because the description is the same in the two documents, this means that the priority document also supports the amended claim.

Now suppose the applicant does claim priority. If the amended claim is not disclosed by the application as filed, it is not disclosed by the priority document, given that in our scenario the descriptions are same in the two documents; but because the descriptions are the same, it also follows that the claim is supported by the priority document. It seems very counter-intuitive that a claim should be properly supported by the application as filed, and yet not entitled to claim priority to a priority document with exactly the same disclosure.

Now change the scenario slightly, and suppose that the claim in question was present in the application as filed. The added matter provision (38.2) is now irrelevant; but by hypothesis, the claim is not disclosed in the priority document even though it is supported by the Canadian application with exactly the same description, which means that priority cannot be claimed.

Refusing a claim to priority on that basis is pointless formalism. The quid pro quo for which the patent exclusivity is granted is the new and non-obvious information disclosed in the description: see Pope Appliance (1929) 46 RPC 23 (JCPC) 55; Halocarbon [1979] 2 SCR 929, 944-45; Pfizer v Ratiopharm 2006 FCA 214 ¶ 23; American Cyanamid Co (Dann's) Patent [1971] 88 RPC 425 (HL) 452; Consolboard [1981] 1 SCR 504, 517; Amazon.com 2011 FCA 328 ¶ 36. The new information disclosed in the specification—the “mental result” or “new idea” or “novel mental concept”—cannot be patented as such: “A disembodied idea is not per se patentable. But it will be patentable if it has a method of practical application”: Shell Oil [1982] 2 SCR 536, 554. The claims define practical embodiments of that information: as the SCC said long ago: “A patentable invention is a mental result. . . . The machine, process or product is but its material reflex and embodiment”: Ball v Crompton Corset Co (1887) 13 SCR 469, 477 quoting with approval Smith v Nicholls, 88 US (21 Wall) 112 (1874) 118. The requirement that the claims are supported by the description ensures that the claimed subject matter reflects the new information: an inventor cannot disclose a new way of peeling potatoes, and claim a cure for cancer. But if the original disclosure was sufficient to support the claim, the patentee has met its part of the bargain, and that is just as true whether the disclosure was made in a foreign filing or in a Canadian application in exactly the same terms. In this scenario, the information content of the two documents is exactly the same; the only difference is that the exact words of the claim do not appear in the priority document. This is an absurd basis for refusing priority—and I mean that in the statutory interpretation sense, that it cannot possibly be right. A claim that is fully supported by the Canadian application should be able to claim priority to a priority document with exactly the same disclosure.

However, this absurdity flows from a fairly straightforward reading of the text of the Act. Of course, while the text is dominant in statutory interpretation, the text must be read in light of context and purpose to arrive at a harmonious result. So, I looked more deeply into the context, including the modern and historical case law, the history of the rule changes, the concept of support, the law in other jurisdictions and so on. That context provided some insights, but added some complexity, and while I have some leads, I have found no easy answer, at least at this point. Unfortunately, I am not in a position to chase this down right now. Instead, I’m just adding this update to say that the position I stated in the post above follows from the text of the Act, but it leads to enough problems down the road that I no longer have the same confidence in it that I did when I wrote the post.

Now consider another implication of the above hypothetical, in which the priority document and the Canadian application have exactly the same description, and the Canadian application has a claim that is supported by the description but is not disclosed by the description. As discussed, that means that the claim is valid but priority cannot be claimed. But at the same time, it is almost certain that a description sufficient to support the claim would also be sufficient to anticipate the claim, with the counter-intuitive result that the priority document would anticipate the claim (but for the grace period), even though priority cannot be claimed. This result is very difficult to reconcile with the view that “disclose” has the same meaning in 28.1 as in 28.2. That is not to say that disclose has different meanings in those two provisions, as the analysis also turns on the meaning of “priority”. The point is only that this is another absurd result flowing from a straightforward reading of the text.

I should emphasize that in the foregoing, I do not mean to endorse any particular test or definition for “disclose” and “support” and “reasonably infer.” So, in agreeing with McDonald J that “disclose” has the same meaning in 28.1 as in 28.2, this was not to endorse the practice of any particular patent office as to the interpretation of that term. I am given to understand that some patent offices interpret “disclose” to require a verbatim or almost verbatim reproduction of the claim in the description. I doubt that is right; I am inclined to prefer the approach taken by Jacob LJ in Unilin Beheer [2004] EWCA Civ 1021 [48].“Disclose” of course is subject to extensive caselaw in the context of anticipation, but even if it has the same meaning in the priority context, it is not obvious how it should actually be applied. For example, a claim to “PDEV inhibitors for the treatment of erectile dysfunction” would be anticipated by the prior use in traditional medicine of epimedium to treat ED; but that does not mean that the disclosure in the priority document of the use of epimedium to treat ED would disclose the use of PDEV inhibitors to treat ED. More broadly, there is very little caselaw on the meaning of any of these terms: see the very interesting discussion of the meaning of “support” in CA/PL 27/99, part of the Travaux préparatoires for the EPC 2000. I expect a proper resolution of this problem will show that some aspects of current practice, or implicit understandings of the key terms, are wrong.

Finally, I note that the result in this case does not seem to have turned on the point. PSETs experts argued that certain claims were “reasonably inferable” from the 976 patent [216]–[217]. McDonald J rejected this argument in terms which seem to me to reject the argument of reasonable inference [218]–[219]. McDonald J did go on to say that the priority document must “disclose” the invention [220]–[221] as discussed above, but, so far as I can tell, this was not necessary to the result. (I want to emphasize that I am not saying that I am of the view that reasonably inferable is the test for claiming priority—simply that the point was not determinative in this case.)

“Disclose” Has the Same Meaning in Claim Date and Novelty Provisions

Paid Search Engine Tools, LLC v Google Canada Corporation 2021 FC 1435 McDonald J

2,415,167 / Paid Search Engine Bid Management

In this decision, McDonald J held all the asserted claims of PSET’s 167 patent were invalid for insufficiency, anticipation, or obviousness [5] in a decision that turned largely on the facts. (If I’m counting right, PSET asserted 64 claims from 6 independent claims.)

The patent relates to a system for search engine bid management. In the early 2000s advertisers were reaching internet users by paying search engines to have their information or advertisements displayed in response to user searches for sites for specified keywords. Advertisers could bid on search keywords, so that a higher bid would get the advertiser a higher place in the search results for that keyword [11]. Advertisers wanted to manage their bids to get the best placement for the least money. This could be cumbersome for an advertiser managing a large number of keywords. The inventors of the 167 patent developed what they called a Keyword Bid Optimizer tool (KBO). The KBO collected information from the search engine on their customers’ keywords and the bids on those keywords, and produced a report showing: the keyword(s), the number of times it was searched (views), the top 20 bids for the keyword, and where the advertiser’s bid (in red) ranked relative to others, as shown in this figure from the patent [17]:

With the report in hand (or on-screen), the advertiser could then easily scan the table and see, for example, that they could reduce their bid on “earth friendly product” from 5¢ to 4¢ while still maintaining their first place ranking. The claimed invention seems to have been an automated version of the KBO tool, which incorporated automatic bid reduction as an essential element [236].

McDonald J held on the facts that two of the key independent claims (1 and 17) were undoubtedly obvious:

[260] This is not a case where the application of the obviousness analysis strains credulity. The evidence is clear that even on a rigid application of the obviousness considerations there were multiple other commercial products that performed the same essential features, and in some cases more, as those described in claims 1 and 17 of the 167 Patent.

This is a classic obviousness scenario, where the development of a new product or technology drives a demand for complementary technologies. There were no doubt many new and useful products that were nonetheless obvious, as companies rushed to pluck the low hanging fruit from the new opportunities opened up by search engine advertising, which was relatively new at the time [62].

A number of the claims were also anticipated. The 167 patent claimed priority from a US provisional patent application, which was referred to as the 976 patent. McDonald J held that a number of the claims were not entitled to claim priority from the 976 patent as the claimed invention was not sufficiently disclosed [222]. PSET argued that the claim to priority is met “if an ‘inference’ can be drawn that the 976 Patent discloses the 167 Patent” [220]. McDonald J rejected this, pointing out that “The word “disclose” appears in various sections of the Patent Act. I do not accept that a different meaning should apply to disclosure as it is used in subsection 28.1(1) such that disclosure is achieved merely if an inference can be drawn” [221]. This holding strikes me as clearly correct. Section 28.1(1), the claim date provision, refers to “a patent disclosing the subject-matter defined by the claim”; s 28.2(1), the novelty provision, states “The subject-matter defined by a claim . . . must not have been disclosed.” It is not just the word “disclosed” that is the same; the two provisions have parallel construction, both dealing with the relationship between disclosure and the claimed subject matter, and of course the claim date defined in s 28.1(1) is incorporated into s 28.2(2). Surely the same meaning was intended. PSET conceded that many of the asserted claims were anticipated if they could not rely on the priority claim [231]. Other claims were anticipated by the prior art [245].

McDonald J also held that a number of the claims were not sufficiently disclosed, particularly those based on automatic reduction of a bid based on the demographic information of searchers, as the patent did not provide sufficient description of how to actually use that information: see eg [199], [204]. In effect, patent says “Check out these cool stats and then optimize.” We don’t see successful insufficiency attacks very often, but this strikes me as a pretty clear case, at least judging from McDonald J’s relatively brief discussion.

Finally, I’ll note that the patentee appears to be a variety of patent assertion entity. PSET did actually implement and sell the KBO tool, but it was not commercially successful, and they stopped selling it in 2003 [20]–[22]. As noted, the invention claimed by the 167 patent is an improved version of that tool, with automated bid adjustment, and it doesn’t seem that PSET ever implemented an embodiment of the invention. So far as I can tell, PSET was one of many firms that jumped into an exploding new market, but they didn’t survive the shake out. This litigation was an (unsuccessful) attempt to monetize the patent.

 UPDATE: See further discussion of the priority point here.

Third Party Prior User Rights

Kobold Corporation v NCS Multistage Inc 2021 FC 1437 Zinn J

            2,919,561 / Bottomhole assembly

As outlined in last week’s post, this motion for summary judgment addresses the new prior user defence set out in s 56. The main operative provision, 56(1) provides that if a person, before the claim date, “committed an act that would otherwise constitute an infringement. . . it is not an infringement. . . if the person commits the same act on or after that claim date.” The key question is what it means for the post-claim date act to be “the same” as the pre-claim date act. Zinn J granted summary judgment on the interpretation of this provision. I’ve discussed aspects of the decision relating to 56(1) in previous posts.

Zinn J also briefly addressed 56(6) and 56(9), which protect third parties. Subsection 56(6) provides that “the use of an article is not an infringement. . . if the article was acquired. . . from a person who, before the claim date . . . in good faith, made or sold. . . an article that is substantially the same as the one used, for that use.” Subsection 56(9) similarly protects a person who uses a service provided by a third party who had provided substantially the same service before the claim date. Note that there is no requirement in these provisions that the third party itself bought the product or service before the claim date; the third party can buy the product after the claim date from a party who sold a substantially similar product before the claim date.

NCS argued that it was protected by this provision because it bought the Blue Bullet element from a third party. The Blue Bullet was part of the packer assembly [27], and presumably an important part; in any event, its exact nature is not important to the decision. Zinn J rejected the s 56(9) argument purely on the facts. There was no evidence that NCS purchased the service from others; on the contrary, NCS was providing the service: [146]. Consequently, on its face s 56(9) does not apply. Zinn J also rejected the 56(6) argument on facts, on the basis that there was no evidence that NCS actually acquired the Blue Bullet from a third party, as is required by the provision: [146].

That turns simply on the facts, but one interesting point concerning the language “substantially” the same, found in 56(6),(9), which is evidently broader than “the same,” which is used in 56(1). Kobold submitted that this was intended to protect a third party “from having to prove what it is using is the same as what was previously available, provided it is still substantially the same and used for the same use” [144]. Presumably this is on the view that the third party may have bought the product after the claim date, and may not have easy access to evidence of what was being sold before the claim date, in contrast to the manufacturer, who should be able to provide evidence of what they themselves were doing before the claim date. That’s at least a plausible suggestion, though I’m not sure that a different substantive test is the best way to deal with that problem. After all, the third party still has to prove that the product is substantially the same as the pre-claim date product, so it still has to have some kind of access to that earlier product, so the evidentiary problem largely remains. If the problem is one of evidence, it might have been better addressed directly, by an evidentiary presumption of some kind. Further, these provisions do not provide a great deal of additional protection to the third party. If the new product purchased by the third party is not “the same” but it is “substantially the same,” then the third party will generally not be able to purchase it anyway; even if the purchaser can take advantage of 56(6) to use the product, it will probably not be able to purchase the product after the claim date, because the vendor who cannot take advantage of 56(1) will likely be enjoined from selling.

In any event, the current Act does have two different standards, “the same” and “substantially the same,” which are evidently different. This case has told us what “the same” means; it will take another case to find out what “substantially the same” means. Kobold’s suggestion is plausible, but not beyond doubt. No doubt we’ll see further speculation as to the purpose of the “substantially the same” requirement, given that it is relevant to a purposive interpretation of the provision.

Zinn J also gave a second reason for rejecting the 56(6) argument:

[149] I also agree with Kobold’s further submission that subsection 56(6) does not apply because the use of the Blue Bullet element is not the potentially infringing act. Claims 1, 9, and 12 of the ‘561 Patent are for the use of tools that incorporate elements that are not present in the Blue Bullet element. Claim 15 is for a pressure equalization tool having elements not present on the Blue Bullet element. It is only the use of the Blue Bullet element in conjunction with the other elements of NCS’s packer assemblies that infringes the claims.

I’m just going to leave this here. Because Zinn J ordered a trial, the facts are not particularly well developed, and I don’t understand why the use of the Blue Bullet element is not the potentially infringing act. It does seems to have been the element that incorporates all of the key elements of the inventive concept. This isn’t to say I disagree; I just don’t understand the point well enough to comment. I am flagging it nonetheless, as it does appear to be some kind of substantive restriction on the availability of 56(6).

Prior User Rights and English Law

Kobold Corporation v NCS Multistage Inc 2021 FC 1437 Zinn J

            2,919,561 / Bottomhole assembly

As outlined in last week’s post, this motion for summary judgment addresses the new prior user defence set out in s 56. The main operative provision, 56(1) provides that if a person, before the claim date, “committed an act that would otherwise constitute an infringement. . . it is not an infringement. . . if the person commits the same act on or after that claim date.” The key question is what it means for the post-claim date act to be “the same” as the pre-claim date act. Zinn J granted summary judgment on the interpretation of this provision.

My last post discussed the central holding, that the subsequent act will be the same so long as any changes do not relate to the inventive concept. In addressing that question I did not address Zinn J’s initial discussion of the meaning of the term “the same,” and whether it should be interpreted to mean “substantially the same act”, as NCS argued [82], or rather “near identicality,” as Kobold argued [88]. Zinn J agreed with Kobold that the more stringent interpretation was correct, even saying that the two acts must be “identical” [92] “in the sense that one can only perform those acts set out in section 42 of the Patent Act which were actually performed pre-patent” [97].

This might seem like a very important point to have omitted; but it is less important than it might seem. Knowing that the test is “identical” rather than “substantially the same” is not particularly helpful unless we know what those terms mean. As I discussed in my last post, it is perfectly clear from Zinn J’s subsequent discussion that when he said “the same” should be interpreted to mean “identical” [92], he did not mean that it must be exactly identical in every respect. He expressly stated that a variety of changes, from paint, to location method, to fracking fluid, would all be permitted. Indeed, after concluding that “the same” meant “identical,” began the next part by saying “[t]his interpretation then leads one to ask if the same act is the manufacture of a device, must it be an identical device that is manufactured to obtain the protection of subsection 56(1)” [99]; he clearly did not consider this question to have been settled by the holding that “the same” means “identical,” and he went on to outline many differences that would be permitted, such as a different paint colour, or the use of a more efficient fracking fluid in the claimed method. So, even though Zinn J did hold that “the same” means “identical” [92], [97], it would a misleading to take that out of context as a summary of his holding.

With that said, I’ll comment briefly on that part of Zinn J’s decision.

His reasons for holding that “the same” means “identical,” were first, “identical” is the dictionary meaning of the word the “same” [93]. Second, “substantially the same” is used in 56(6),(9) and “the same” must mean something different, and presumably more restrictive [94]. Third, the phrase “the act” in the corresponding provision of the UK Act, s 64, had been interpreted as meaning “substantially the same” and to the extent the Canadian drafters were looking to the UK Act, the decision to use “the same act” instead of simply “the act” implies they were looking for a more restrictive meaning than under the UK Act [95], [79]. Fourth, even if an identical act is required, this is still more expansive protection than under the previous provision, which restricted protection to the same specific article [98].

The most significant point is the third, as it implies that the test for similarity in Canadian law is more stringent than in UK law, which in turn means that UK caselaw will be of limited assistance at most. I am not persuaded by this point. Now, I should say that this doesn’t make much difference, as we’re not going to find much guidance in the UK cases anyway. So far there are very few UK cases discussing s 64, and none of them go into what it means to be the same acts, and they mostly turn on whether effective preparations were made. But the point perhaps worth exploring nonetheless, in case future UK decisions are more helpful.

The UK Act says a person who does “an act” before the claim date has the right “to continue to do the act.” “The” is a definite article, which, at least according to Wikipedia “denot[es] . . . things already mentioned, under discussion, implied or otherwise presumed familiar to. . . readers.” So “the act” refers to an act already mentioned, namely the prior act referred to in the first part of the provision. So I would say that as a matter of grammar “the act” in the UK provision means the same act as was previously mentioned in the sentence, which is to say the prior act. So I have a hard time giving much significance to the inclusion of the word “same” in the Canadian provision, as it seems to me to have been grammatically implicit in the UK Act.

More importantly, “the same” was interpreted as meaning “substantially the same by Jacob J in Lubrizol v Esso [1997] 114 RPC 195 (Pat)*, in a passage quoted by Zinn J at [81] (original italics, underling added by Zinn J):

In deciding whether the activity is substantially the same all the circumstances must be considered. Both technical and commercial matters must be taken into account. That is important in a case such as the present where there are inherent minor variations in starting materials or the like. If the protected act has to be exactly the same (whatever that may mean) as the prior act then the protection given by the section would be illusory. The section is intended to give a practical protection to enable a man to continue doing what in substance he was doing before.

I don’t see any real between what Jacob J was saying in this passage and Zinn J’s holding in the remainder of the decision. As noted, Zinn J explicitly held that changes in paint colour, location method, and fracking fluid, would all be permitted, even though the latter two are actually elements of the claim. This goes well beyond the example of “inherent minor variations in starting materials” given by Jacob J. So when Jacob J held that the act had to be “substantially” similar, this was by way of rejecting a requirement that the acts by strictly identical—a requirement which Zinn J also rejected, notwithstanding that he did say that “the same” means “identical” at [92]. I also note that at [97] Zinn J explained “identical” as meaning “in the sense that one can only perform those acts set out in section 42 of the Patent Act which were actually performed pre-patent” [97] and that “[o]n this interpretation [ie identical] if the person claiming the protection of subsection 56(1) had, for example, previously only manufactured and used its device, then they can continue to manufacture and use it, but they cannot now sell the device to others to use” [98]. This suggests that “identical act” means the type of infringement—whether it be the make, sell, or use—that must be identical.

On the whole, I don’t find this discussion of whether “the same” means “identical” or “substantially” the same to be helpful. The label doesn’t matter. The real question the nature and extent of the differences that are permitted. Zinn J’s lengthy discussion of that central question, discussed in Monday’s post, surely supersedes his statement that “the same” means “identical.”

*Zinn J cited the passage as “[1997] RPC 195 (Eng CA) at 216,” which is a pinpoint to the decision of Jacob J in the Patents Court. Aldous LJ in the EWCA did quote and affirm the last two sentences of that passage, [1998] 115 RPC 727 (CA) 700, but did not quote the part related to “substantially the same”.