IGT / Acres Gaming Inc, Re

When a claim defines an invention partly by reference to excluded subject-matter e.g. a business method, how do you search the prior art? This appeal illustrates some of the difficulties.

Patent application GB 0311200 (Footnote: ¹)was filed by Acres Gaming Inc (Footnote: ²), of Las Vegas, Nevada. It states the basic problem with disarming candour – “to keep people coming back despite losing money”.

It then describes a known method:-

So even casinos are issuing loyalty cards nowadays – a sort of frequent gambler card. The gambler puts his card in a slot where it is read electronically. The data is sent to a central computer for analysis. His identity is ascertained as part of that analysis and so he establishes his entitlement to his bonus points.

The patent application explains that issuing one of those loyalty cards is a rather cumbersome process. The new customer’s data (name, address, etc) has to be keyed in manually. It also takes time to make the card and post it to him. What is more, returning customers sometimes forget to bring their loyalty card.

To overcome those difficulties, the applicants came up with this general idea: Why bother to issue cards in the first place? Why not use somebody else’s cards instead?

By “somebody else’s cards”, I mean the sort of conventional cards – not issued by casinos – that people are likely to carry around with them anyway: for example, credit cards, supermarket loyalty cards, or (in America) driving licences, or social security cards. The information may be embedded in a magnetic strip or a chip attached to the card or (in the case of an American driving licence), a barcode.

As described in the patent application, what happens at the casino is this. The customer takes one of those cards – his credit card, for example – and presents it to a reader. The binary number stored in the card is read electronically and is sent to a computer, which looks in a database to see if the customer already has an “account”. If he has, he gets the privileges that pertain to that account: bonus points, or whatever. If he has no account he is invited to open one.

The patent application describes a further idea: that, in order to search the database, the computer does not need to decode the raw information that is read off the card. As it explains:-

Therefore the system can handle a variety of cards whose encoding protocols may be mutually incompatible. To search the database it just uses the raw number read off the card. The raw number may encode useful information about the returning customer e.g. his name, but this can be ignored (at that stage, anyway).

Claim 1 of the patent application as originally filed read as follows:

What is defined in that claim differs from loyalty schemes already in use in casinos, and admittedly so, in one material respect only: that the card is a pre-existing card issued by an entity other than a casino. But that is just an institutional fact, denoting a legal relationship. It may or may not have been a clever business idea to piggy-back on pre-existing conventional cards, but what is there defined is not a patentable invention according to the law of this country. The innovation lies in the world of business administration, not technology: see below.

Furthermore, the concept was not new even in terms of casino business methods. It had already been proposed – see US Patent 5811772A (Lucero) – that in order to obtain playing credit without having to queue up at a window to get more coins, a player could feed a general purpose charge card (e.g. Visa, MasterCard, American Express) to a reader at a gaming machine. But Lucero also disclosed, see Figure 8, that the information read off that general purpose charge card could be used to ascertain if the player already had an account with the casino. Of course, the general purpose charge card would be a card “issued by an entity other than a casino”.

Therefore, Claim 1 as originally filed could not be allowed and the applicants accepted it. So they proposed an amended version.

Claim 1 of the UK patent application as presently on file reads as follows:-

That was said to be based on the passage I have quoted in paragraph 9 above. From certain subsidiary claims in the application (Footnote: ³) it appears that decryption of the information at a later stage for a different purpose (i.e. after the account has been identified and accessed) is not ruled out.

There are other claims on file, but having considered these I believe I can decide this appeal essentially on the basis of existing Claim 1.

Article 52 of the European Patent Convention reads as follows:-

Article 52 has proved to be difficult to apply but the approach that is current in this country was laid down by the Court of Appeal in Aerotel Ltd v. Telco Holdings Ltd [2006] EWCA Civ 1371. There are four consecutive steps:-

The fourth or last step is just a cross-check, because “technical” is too vague a test to be used on its own.

In discussing the second step the Court of Appeal said this:-

The Court of Appeal accepted those submissions and I have stressed the most important point: what has the applicant really added to human knowledge?

After hearing argument in this case I wondered what is meant by the second paragraph I have quoted, namely paragraph 44. Does it mean that the Patent Office is bound to accept the applicant’s assertion (save in blatant cases)? Or can the Patent Office do a prior art search to find out what has the inventor really added to human knowledge? I therefore invited further submissions in writing.

Although there was some disagreement, both parties accepted that the Patent Office is entitled to do a prior art search and that if it turns out that the alleged contribution was already known, or was obvious, there can hardly be a contribution to human knowledge. In my judgment that is correct. And there will be no patentable contribution to human knowledge if what is new and not obvious relates solely to a business method as such.

The Examiner (Mr Paul Marshall) objected that Claim 1 as amended still did not define a patentable invention because it was for a business method or computer program as such. He also objected that what was claimed did not involve any inventive step. There was a hearing at which Mr Thomas Moody-Stuart represented the applicants, as he does now.

In his decision the hearing officer (Mr R C Kennell) first considered whether the claims of the application defined a patentable invention. He proceeded as follows. He cited the four-stage test laid down by the Court of Appeal in Aerotel. He assessed the inventor’s ‘contribution’ (i.e. what he had added to human knowledge) as follows: “The use of encrypted information on a pre-existing card, without decrypting it, for the purpose of either identifying an existing account or confirming whether or not an account already exists”. However he then went on to consider whether the ‘contribution’ related solely to excluded matter. He held that it did because it was “as a whole, sufficiently part of the administration of the business to be excluded as a method for doing business” irrespective of whether the invention was claimed as a method or as apparatus. He therefore held that the application should be refused.

Although the hearing officer thought it was not strictly necessary for his decision, he then went on to consider whether the claimed invention was obvious; but he was not satisfied that it was. The inventive step could be the recognition that there was information in a pre-existing card which could be used as an identifier without decryption, “which would not have occurred to the skilled man without inventive effort”. (I should explain that the Examiner had argued this point without relying on any specific piece of prior art. His argument was that it was well known to use a string of identifying information in order to access a computer record, and that lack of ‘decryption’ was a red herring.)

In opening this appeal Mr Moody-Stuart pointed out that, if indeed it was not obvious to use the identifier on a pre-existing card without decryption – as the hearing officer was prepared to hold – then the invention of amended Claim 1 ought to have been held to be patentable. For in that case the contribution that the inventor had made to human knowledge did not consist solely of excluded matter.

So far as that goes, I would have been prepared to hold that Mr Moody-Stuart was right. If – and I stress the word “if” – it was indeed new and not obvious to those skilled in the art at the priority date of the patent application to devise an apparatus where a card bearing encrypted information is presented to a reader and the information is used to access a record in a database without decrypting the information first, it would be so irrespective of the business (if any) in which the concept was to be employed. It would be a new and inventive technical procedure. It could therefore hardly be said to be subject-matter falling solely within the excluded province.

Furthermore, even if Claim 1 failed, other claims might not, and the applicants would at least be entitled to those. Claim 11 as now on file is for an apparatus according to Claim 1 where the pre-existing card “is drawn from the set including a driver’s licence, a credit card, and a bonus card”. Such cards have mutually incompatible encoding protocols. So at that level of generalisation the invention could be described as follows: In an environment where cards bearing information encrypted according to mutually incompatible protocols may be presented to a card reader, the use of the raw encrypted number to access a record in a database without decrypting the number first. I have stated the concept without reference to any business method as such, and if that concept is indeed new and not obvious, the applicants would be entitled to patent it.

Although there are some claims (Footnote: ⁴) of the application which are directed to a computer program which, in use, would cause the above-described procedures to be performed, it would seem that those would not be considered to be computer programs ‘as such’, in the assumed premises and having regard to the recent decision of Kitchin J in Astron Clinica Ltd v. Comptroller-General of Patents [2008] EWHC 85 (Pat).

Therefore, if that had been all, I would have been prepared to allow the appeal and to direct that the patent application should be granted.

However, for reasons that emerged during the hearing and which I shall now describe, I am not satisfied that what I have said so far is a completely accurate description of the case.

Claim 1 as now on file uses the expressions “encrypted”, “non-decrypted”, and “without decrypting the information”: see paragraph 15 above. In my judgment they lack sufficient clarity and may have misled the Examiner when he searched the prior art, so that it is possible that there existed prior art in the light of which the concepts I have described were not new, or were obvious

The data in these cards may be stored in a magnetic strip, a microchip, or the like. If you could somehow read one of those with your own eyes, you would see a string of bits – a sequence of 1s and 0s. They constitute a binary number that encodes certain information pertinent to the holder of the card. In some instances there is nothing secret about the method of encoding. For example, American driving licences encode the licence number, the issuing State, the driver’s name, address, and so forth, according to an agreed and published encoding protocol. Anybody skilled in the art could recover that information using standard methods.

In other instances, however, the information (or some of it) is intended to be kept secret. For example, if your credit or debit card is stolen it is important to stop thieves recovering your PIN. The correct name for that kind of encoding is enciphering. Enciphering is a species of encoding.

Unfortunately, the claims of the patent application, as now on file, use a third expression: encrypted. I cannot find that that word was used in the application as originally filed, and in my judgment it is unduly obscure as it stands. It could suggest that some of the information in the card is meant to be kept secret, which is indeed the primary meaning of the word. It is true that if you read the document as a whole, carefully, you may be led to think that that is not the case, because of the references to driving licences and social security cards. I believe that in America such documents do not encode secret information but, as against that, this is a patent application for the United Kingdom. Therefore, how can the skilled reader be sure that the applicants did not decide to disclaim the use of cards bearing nothing but open-access information?

In response to questions I put to him at the hearing, Mr Moody-Stuart told me that Claim 1 was intended to cover both cases in the alternative i.e. (1) where at least some of the encoded information is secret or (2) where none of it is.

When I first perused the document I thought that “encrypted”, and so forth, referred to enciphering. Furthermore, Mr Colin Birss, who represents the Patent Office, told me that the use of encrypted, and so forth, had misled the Office during the official examination, perhaps causing them to miss relevant prior art. They asked me for a further opportunity to examine this application.

I was not told why the original expressions “decode”, “encoding”, “encoded”, “without decoding”, and so forth, were changed. If e.g. “without decrypting” means exactly the same as “without decoding”, why change it? If it does not mean exactly the same, is there added subject-matter? Maybe an intermediate generalisation was intended. I do not attempt to answer those questions now. It is enough that the expressions currently in Claim 1 are not sufficiently clear.

A patent claim should desirably be as clear “as the subject admits of” (Footnote: ⁵). It may be possible to arrive at the most probable meaning of a claim by paying careful attention to the text of the document as a whole, but at the examination stage this many not be enough. A higher standard can be required. It should be remembered that, in real life, industry may have to consider a large number of patents and in practice there may not be enough time to do more than look at the broadest claim and the drawings. If the claim seems to bear a meaning that is, in fact, inconsistent with its real meaning when read in the context of the whole document, the reader may be misled. The claim is therefore obscure, and if the obscurity is avoidable objection arises under section 14(5)(b) of the Patents Act 1977.

In my judgment the Patent Office is entitled to take the point even at this stage and I have power, in the exercise of my discretion, to allow them to do so: see section 99 of the Act. Although I bear in mind that the point has been taken very late in the day, I exercise my discretion having regard to the public interest i.e. that it is most desirable that patent claims should be clear. Not just that, but the law so requires: see section 14(5)(b). I therefore hold that this patent application needs amendment for the reason I have stated. Although the time has expired for the case to be in order for acceptance, the case is before me under appeal, and the combined effect of the Act and Rules is to give the applicants an opportunity to amend. By the same token the Patent Office is entitled to examine the document as sought to be amended, and to search for further prior art if so desired.

I shall therefore put Mr Moody-Stuart’s clients to their election. They can either accept the consequences I have stated in the preceding paragraph. Or they can choose to stick to the language they have chosen, apply to the Court of Appeal for permission to appeal, and seek to justify it before that tribunal – I do not encourage this myself. (They will also run the risk that their patent, if granted, could be held invalid for added subject-matter.)

That is enough to dispose of this appeal, but I should like to add some brief remarks about searching the prior art in these cases in case it may assist the Patent Office in future. (I need hardly stress that no criticism whatever of the search examiner is intended.)

In CFPH’s Application [2005] EWHC 1589 (Pat) I pointed out that “patents that are wrongly granted can be very expensive to challenge, and perhaps beyond the means or inclination of small and medium enterprises. An accumulation of patents of that sort (sometimes known as a "patent thicket") may be a serious barrier to entry. The only safeguard against that wrong – and it is a wrong – is the vigilance of the Patent Office.” That is consistent with what the Court of Appeal said in Aerotel at paragraphs 17-21.

It is therefore in the public interest that patent applications should be searched against the prior art with appropriate thoroughness, and that is what the Patent Office normally tries to do.

Patent applications are usually searched against an existing patent database, but because business methods as such are not supposed to be patented and should not be in the database in the first place, relevant prior art could be missed if the search were to be confined to subject-matter partly defined by mere “business” features. The most relevant prior art might be a document propounded in a wholly different business context, or in no business context at all. It is therefore legitimate, when performing a search, to strip out the “business” features and to search for the remaining combination of “technical” features.

For instance (and to take the present case as an illustration), the search need not be limited with reference to casinos and “accounts”. Such features can be stripped out for the purpose of the search. For example, the subject of existing Claim 1, when suitably clarified, might be searched against the following concept: “Reading the information encoded in a card or other token and using the raw number thus read to search a database without decoding the number first” (Footnote: ⁶). It would still remain for the substantive examiner to decide whether the claim is for patentable subject-matter having regard to the prior art thrown up by the search, and according to the principles laid down in Aerotel.

Furthermore, when it is suspected that the sole contribution to human knowledge may be a business method as such – i.e. the “technical” features of the claim were known or obvious – it does not follow that a patent database is the only or best place to search. The most relevant prior art document may not be in any patent database, because it was not decided to apply for patent protection. This could be so where the document is a published international standard. Such documents may be available online.

Purely by way of example, passports have a machine-readable string at the end. Likewise an American driving licence is machine-readable, and identity cards have been proposed which would be machine-readable. Also there could be traveller’s cheques with a unique identifying code in a magnetic strip (see e.g. WO9724694). If such a document is reported as stolen (or if the authorities have some other reason to interview the bearer) the fact may be flagged up in a database. Is it known to search the database without decoding the string first? I do not know; but the Patent Office is entitled to find out if a concept of that sort has been published. That is a matter for them.

The point I am venturing to make, therefore, is that when the Patent Office wants to do a search it is entitled to strip out the merely “business” features of a claim. This may be known to examiners already, but I do not believe it is stressed in the Manual of Office Practice.

Furthermore, if the stripping-out exercise is performed it may become self-evident that the combination of “technical” features was common general knowledge, or even acknowledged in the application itself (see e.g. paragraph 12 above). It may then emerge that a search under section 17 would serve no useful purpose, see section 17(5), and may be deferred pending amendment, thus saving valuable human resources.

But on matters of searching I do not want to say one word that would fetter the discretion of the Patent Office in this and future cases.

Because of the outcome of this appeal there should be no order as to costs.