The Patent Office has properly started citing the Bilski decision to support rejections of process claims for failure to define patentable subject matter. This LA Times article notes example Bilski-based rejections of claims for processes ranging from seismic data analyses to methods for converting an internet domain names to read both left to right, for languages like English, and in the opposite direction, for languages like Arabic and Hebrew.

Such rejections are expected, of course – the Patent Office is simply applying the current law, as it should. But a recent patent granted to Chase Card Services, the credit card division of JP Morgan Chase, has to make you wonder how strenuously the Office is applying the new test.

United States Patent No. 7,480,631 is directed to fraud detection processes and was granted on January 20, 2009. Claim 1 is reproduced below:

1. A method for detecting and processing fraud and credit abuse, the method comprising: receiving a telephone call from a caller concerning a credit account; identifying one or more potential fraud-related problems by searching the credit account data for conditions that match one or more predetermined qualifiers and quantifiers; presenting the caller with a series of questions and soliciting, from the caller, responses to the series of questions, wherein each question in the series is generated based on the one or more potential fraud-related problems and further based on the caller’s responses to earlier questions in the series; and determining a fraud status associated with the credit account based on a computerized analysis of the solicited responses.

Where’s the Bilski-required machine, apparatus, or transformation in that claim? Is it the ‘computerized analysis’? If so, is the term “computerized” sufficient to tie this claim to a particular computing machine? Beyond that, I can only wonder, tongue only partially in cheek, whether the ‘telephone’ provides the machine or apparatus.

It’s interesting to compare the Patent Office application of Bilski, which produced the claim above, to the approach taken by the Federal Circuit. Comparing this claim to the claim summarily rejected by the Federal Circuit in its post-Bilski decision in Classen Immunotherapies v. Biogen IDEC, which included a step of “immunizing mammals,” suggests that the new ‘machine, apparatus, or transformation’ test is not being applied consistently between the Office and the court. On one hand, the Office appears willing to find a required connection to a particular machine or apparatus either in a vague term like ‘computerized’ or through the use of a known apparatus like ‘telephone.’ On the other hand, the court appears completely unwilling to recognize a step that recites a physical act that is well-known to trigger a physiological change as a required transformation.

The claim of the ‘631 patent is, of course, anecdotal in nature. It represents only a single post-Bilski patent issued by the Office and surely there are numerous examples of applications in which examiners took a strict approach to applying Bilski. But the existence of even one issued claim like the one above is troublesome because it reveals the difficulty associated with applying the vague test against which all process claims now must be evaluated.

UPDATE: A quick glance at the prosecution history for the ‘631 patent shows that the examiner mailed the Notice of Allowance on October 17, 2008 – 13 days before the court announced the Bilski decision. The timing, of course, doesn’t change the impact of the decision on the claim or render it subject to a different test for patentable subject matter. It may, though, explain the absence of a clear tie to a machine, apparatus or transformation – while the patent didn’t issue until January, the administrative gears of the issue process had already been set into motion and, apparently, could not be stopped.

In my mind, this actually raises more questions than it answers. Should the Office be required to evaluate all allowed claims that have yet to issue as patents when an intervening major change in patentability law is announced? What about a minor change? These are, of course, administrative headaches and would be time-consuming and expensive to implement. Should the patent attorney be responsible? Should prosecuting attorneys be required to amend claims or present additional arguments when such an intervening change is implicated? How about when such an intervening change might be implicated?

I wonder if this explains why Chase waited nearly 5 weeks to announce the granting of this patent.

Understanding the timing of these two cases is important. On October 30, 2008, the court announced its much-anticipated decision in In re Bilski in which the court adopted the machine-or-transformation test for patentability of processes under 35 U.S.C. 101. The court heard oral arguments for Classen after oral argument for Bilski but before publication of the Bilski opinion. This timing means that members of the Classen panel were likely debating the Bilski issue and/or reviewing the Bilski opinion at the time of the Classen oral argument (The court sat en banc for Bilski).

Indeed, listening to the Classen oral argument clearly shows that Judge Moore was intensely interested in the patentability of mental processes at the time. She initiated a discussion on the issue by asking counsel for the patentee to ignore the non-mental immunizing step of the claim at issue (“yes, clearly you’re absolutely right…these are not exclusively mental process claims.”). Counsel entered into an academic discussion with Judge Moore only reluctantly and eventually offered an interesting perspective on why such claims should be patentable (just because you can’t enforce a patent doesn’t mean you should be prevented from getting the patent).

Then, after this fascinating discussion, the court issued its one paragraph opinion that summarily dismisses the claims as unpatentable. To support this holding, the court pointed only to Bilski. This would be unremarkable, of course, if the claims at issue in Classen were pure mental process claims. But, as is clear from the oral argument, the claims included a non-mental step (immunizing). Indeed, as mentioned above, Judge Moore had to hypothetically remove this step from the claim in order to entice counsel into the patentability discussion in the first place.

So what became of the non-mental immunizing step? We really have no idea. But, considering the lack of discussion in the Classen opinion and its non-precedential status, we know that an open question exists as to whether a non-mental step (particularly, a known non-mental step) in a process claim that can otherwise be performed mentally is sufficient to confer patentability.

How’s that for an interesting one paragraph opinion?