Anticipation continues to mount as commentators anxiously await the handing down by the US Supreme Court of it’s decision in Bilski v. Kappos, heard in November 2009.
There has been longstanding controversy and vigorous debate over the extent to which software patents and computer related inventions should be eligible to receive patent protection. Those in favour of such patents being awarded argue that they promote investment in research and development.
Those campaigning against the issuing of software patents point to the exorbitant costs of determining whether a particular piece of software developed infringes any claims in existing software patents. They say even if a small software company or an individual inventor engages a patent attorney to provide a clearance search, there is still
no guarantee that it will be thorough enough to prevent them from becoming embroiled in
expensive infringement proceedings.
Developers faced with patent infringement notices by those in possession of huge software patent portfolios say they can’t afford to pay what patent holders demand in licensing fees to ward off infringement proceedings. They also point to the huge volume of patents being granted over relatively trivial advances. Whilst it is always open to developers to challenge the validity of patents granted, either by re-examination or opposition proceedings the cost is often prohibitive. Software patents affect smaller enterprises that don’t possess a large defensive patent portfolio.
Sometime this month, probably towards the end of June, the US Supreme Court on a petition for a writ of Certiorari from the Federal Circuit Court, will deliver it’s opinion as to
whether the Federal Circuit Court of Appeals (CAFC) erred in law in concluding that a ‘process patent’ must be tied to a machine or apparatus or transform an article into a different state or thing to be eligible for patent protection. This test is found in previous case law and has become known as the ‘machine or transformation test‘. The US Supreme Court quashed the CAFC decision in Bilski v Kappos, agreeing that the law in relation to the patentability of method patents was in need of review and clarification.
The legal saga commenced when Bernard Bilski and Rand Warsaw sought a patent in 1997 over a method for predicting and hedging risk in commodities trading. The patent examiner rejected all of the eleven claims set out in the Patent Specification filed by
the Applicants, on the grounds that the “the invention is not implemented on a specific
apparatus and merely manipulates an abstract idea and solves a purely mathematical problem without any limitation to a practical application, therefore, the invention is not directed to the technological arts.”
On an appeal to the Board it was held that the Applicants’ claims did not involve any
patent-eligible transformation, and the transformation of merely “non-physical financial
risks and legal liabilities of the commodity provider, the consumer, and market
participants” can’t be said to constitute patentable subject matter. The Board also
concluded the patent did not provide a “useful tangible and concrete result”. From the
reasoning of both the Board and the Patent Examiner it was apparent that there was a
divergence of opinion as to the appropriate tests to be applied to business method or
process claims. It is clear that part of the reasoning was based on the application of
the ‘machine-transformation test’
An appeal to the CAFC was unsuccessful. The Federal Circuit, after surveying the case law in this area, came to the conclusion that a process is eligible for patent protection if it is (a) tied to a particular machine or apparatus or (b) it transforms a particular article into a different state or thing. Bilski’s patent, being a business or process patent, wasn’t “tied to a machine or didn’t transform an article”.
This landmark decision signifies a shift away from previous Court rulings in which patents
for intangible business processes and methods have been granted protection. The CAFC
decision in Bilski has led to the refusal of more business method patents by the United
States Patent and Trademarks Office (USPTO).
The question is whether this means that patent eligible inventions will be confined in the future, through the US Supreme Court ruling, to tangible objects tied to a machine, rendering any pending and existing patents of questionable status, where they don’t employ either a machine or a man-made thing.
Many hardware manufacturers, high tech companies, software houses, telecommunication and biotech companies await the outcome of the case to gain a clearer understanding of the
boundaries of patentable subject matter in relation to business method patents, and the extent to which they will be able to patent software, financial strategies and other abstract
The case has commercial implications for the patentability of business methods
and software and those operating in the field of biotechnology, financial services and
information management systems. Software developers derive the bulk of their revenue through licensing their inventions for third parties to use, particularly with regard to inventions of general application, such as innovative compression, encryption, storage techniques or processes.
Software per se is not regarded as patentable subject matter as patents exclude alogirithms, being merely abstract mathematical concepts or ideas. However, the previous Supreme Court decision in Diamond v Diehr had created a legal foundation for permitting a patent to be granted over a physical process that is controlled by software.
Although legal commentators have questioned the validity of the Diehr decision, which held that “anything under the sun can be patented”, the Supreme Court of the United States has not had an opportunity since the Diehr case to review the patentability of software. The Supreme Court is now presented with that opportunity to revisit the Diehr and other authorities, and define with more precision the contours of software patent protection.
The US Supreme Court’s decision comes at an important time, given the explosion of applications for protection in recent years for new methods of doing things using new technologies. The internet has given rise to many novel business approaches. The fact that the Supreme Court even accepted the petition has been greeted as a sign for optimism particularly in light of the fact that the Government had been urging the Court against taking this path.
However, it would be an oversimplification to suggest the Bilski ruling will deem all
software related patents either invalid or unpatentable. Whilst there are other known forms of intellectual property protection for software such as copyright, those advocating the granting of patent protection for computer programs contend that there are inherent
limitations associated with trying to protect the intellectual property via copyright.
The source code represents protected expression under traditional copyright law, however it is relatively easy for programmers to use different programs to alter code whilst preserving it’s functionality, which copyright is not intended to protect. The format in which software inventions are disclosed in patents, namely using plain language, line drawings and flow chart diagrams enables any third party with programming skills to reverse engineer software thereby fulfilling the underlying ideas the subject of a patent. There is a notable trend in the U.S. for relying upon patent protection to protect software and computer programs, noting how copyright law has narrowed the protection conferred upon software to it’s intended limits. Computer Associates International Inc. v. Altai Inc.
The debate over the preferred scope for copyright protection has been a contentious one and software has become more ubiquitous, assuming more commercial and functional importance than hardware.
There has been a proliferation of applications for patent protection filed around the world,
particularly in the US. The controversy over the patentability of software applications is due in part to particular features associated with the patent process such as disclosure
requirements, examination standards and the limited nature of prior art material available to be examined by the Patent Office. This inhibits to a Patent examiners’ ability to survey the prior art base in the assessment process.
The subject matter of the Bilski case is not simply the eligibility of software for patentability. It is important to remember that the particular subject matter upon which the US Supreme Court will be ruling in Bilski is a business method patent, not a software patent. It is a method of commodities trading proposed by the Applicant for which the USPTO refused to grant a patent. The Applicant Bilksi appealed to the CAFC, which rejected the patent because the idea didn’t use any particular machine and didn’t exhibit any transformation to satisfy the test for patentability.
When the Supreme Court delivers it’s judgement it may even confine itself to focusing on
the patentability of business methods, or alternatively it may render a broader decision
which may or may not affect the patentability of some categories of software.
The real significance of the case for software patents is that some software patents
customarily granted by the USPTO would also fail to meet the requirements of the “machine or transformation test“. Furthermore, there is precedent in 2009 for courts rejecting the eligibility of certain software patents on this very basis. In reviewing the decision of the CAFC, the US Supreme Court will deliver an opinion on whether the Bilski method patent should be upheld or rejected. If the Court finds that it should be refused protection, it is in a position to either re-affirm the CAFC’s machine or transformation test, or enunciate a new test. This test could potentially invalidate a broader or narrower subset of patents, including software patents.
Even if the Court adopts and re-affirms the machine or transformation test, whilst this could invalidate a particular set of software related patents, it would still leave intact and
unaffected the validity of other categories of software patents. The outcome will depend
upon whether the Supreme Court decides to take it upon itself to confront the question of
the legitimacy of software as a category of eligible patent subject matter. The machine or transformation test, as applied to process patents, means that such a patent must be either tied to a particular machine or apparatus or must operate to change products or materials into a “different state or thing”. Even under this test, some process or method patents wouldn’t be totally disqualified.
Of course it is still open to the US Supreme Court to adopt a test which ignores software
altogether, effectively reversing the CAFC’s Bilski court’s finding. Just as re-affirming
the CAFC’s ruling wouldn’t necessarily disqualify all software patents, if the court chooses
to ignore the CAFC’s ruling, this wouldn’t automatically mean that all software patents
would be patentable. If the Court were to reject the CAFC’s test for patentability, that
doesn’t mean the Court would be rejecting the substance of the CAFC’s conclusion. The US Supreme Court may decide that the CAFC merely misapplied the test in the case to the application before it.
Many commentators predict that the Supreme Court will use the Bilski case to
enunciate a broader test which will impact on the patentability of software, perceiving
business method or process patents as part of a similar category as computer programs. A
computer program is just a sequence of instructions which can be executed by human beings or computers. The Circuit Court’s ruling is quite broad, therefore it seems unlikely that the Supreme Court will uphold the Bilksi patent and adopt a test which renders the scope of patents even broader. The more likely scenario is that the decision may be scaled back slightly and narrowed in it’s application. Considering the case law and the various tests which are found within it, the US Supreme Court could elect to adopt any one of them.
However, contrary to what has been published, the Supreme Court is unlikely to issue a ruling which merely declares all software patents as invalid. Rather, the more likely result is that they will adopt a test which confers greater or lesser discretion to reject software patents. A realistic outcome would see the US Supreme Court fail to uphold Bilski’s method
patent, without necessarily discarding the machine or transformation test. The US Supreme Court can still declare that it be retained as one of a range of tests to be applied in considering the eligibility of any given patent.
Some have argued that the Bilski ruling will present obstacles to the future patentability of medical diagnostic devices and treatments, on the basis that you cannot treat something which isn’t first discovered. However along the lines argued above, this still doesn’t preclude such innovations being found to be patentable, as the case of Arrhythmia Research Technology, Inc. v Corazonix Corp demonstrated. The relevant invention
consisted of a process for analysing electrocardiograph signals directed towards determining whether a patient is predisposed to ventricular techycardia, a form of cardiac arrythmia.
All of these processing steps were able to be executed using a digital computer with appropriate software connected to an electrocardiograph machine. The basis of the District Court’s objection was that the claims were directed to a mathematical algorithm. The CAFC held in that case that a patent application directed to analysing heart signals to be patentable subject matter as the invention transformed the signals themselves from analog to digital form. In arriving at their decision the Court applied the ‘transformation’ limb of the machine or transformation test. It was accepted that the method involved a physical process with steps that transform one physical, electrical signal into another. Arrhythmia provides support for the contention that a claim to a software invention need only recite a new process or apparatus.
The Supreme Court is unlikely to issue a ruling which merely declares all software
patents as invalid. Rather, the more likely result is that they will adopt a test which
confers greater or lesser discretion to reject software patents.
A realistic outcome would be that the US Supreme Court will fail to uphold Bilski’s method
patent, and whilst it won’t discard the machine or transformation test, will declare that it
should be retained as one of many tests to be applied in considering the eligibility of any
given patent. The Bilski ruling will present obstacles to the future patentability of
medical diagnostic devices and treatments, as you can’t treat something which you don’t
first discover. However along the lines argued above, this doesn’t preclude such
innovations being held patentable under other tests as the case of Arrhythmia demonstrates.
PATENTABILITY OF SOFTWARE AND METHOD PATENTS IN AUSTRALIA
As stated above, source code of computer programs isn’t, in and of itself, protected by the patent system, and can be protected by either copyright or trade secrets. However software is patentable where it describes methods of operating computer systems or other computational methods which result in a commercially useful result. Neither the source or object (machine code) is required to be disclosed in a patent specification. Both ecommerce software applications and business method patents have been granted by the Patent Office for some time.
Applications for software patents are treated by the Patent Office in the same manner as any other application for an invention, namely they must satisfy the same tests of meeting the definition of patentable subject matter and passing the threshold tests of novelty and inventive step. Patentable subject matter in Australia can be broadly defined as a useful man-made process or product as distinguished from an abstract idea, mathematical algorithm or laws of nature.
IBM was granted a patent for an invention related to a computer program rendering an
improved curve image on a visual display by the Federal Court of Australia in 1992 ( see
International Business Machines Corporation v. Commissioner of Patents). In considering whether IBM’s claim constituted patentable subject matter, the Court cited the seminal decision in National Research Development Corporation v Commissioner of Patents, as representative of the leading authority in Australia (the NRDC decision). The High Court in NRDC had required that for a process to be eligible for patent protection it must be one that offers some material and useful advantage belonging to the useful arts as opposed to the fine arts. The improved curved image was found to be a commercially useful effect in computer graphics, and not merely a claim reciting a mathematical algorithm.
Patent office guidelines and the Federal Court of Australia have subsequently accepted that if software had a commercially useful effect, then it could be considered eligible for patent protection. A commercially useful application can only arise when a mathematical algorithm is implemented in some way so as to produce a result. This appears to be a broad test as most software applications do have some commercially useful application.
The Full Court of the Federal Court of Australia seemed to be suggesting that the scheme
would need to be embodied in some kind of software or processing system to have some
physical effect or application to be patentable. This reasoning wasn’t well received, and
has since given rise to debate as to what type of “physical effect” would be necessary to
render such a method or process patentable. The Court was prepared to state that a “physical effect” would encompass the ‘transformation of data’, ‘an application of a computer program, the operation of a method in a physical device, or a change in state or information in a part of a machine.
The most recent High Court decision in the area of method or process patents was Grant v The Commissioner of Patents, in which the invention in question, involved a method of protecting assets from financial risk in the event of a finding of legal liability. The invention, as filed, didn’t allude to any technical features, data processing or computer system. The High Court held that the patent claimed by Grant wasn’t within the scope of patentable subject matter, and that commercial and financial schemes, like abstract principles of nature are the patentable.
IBM has recently commenced marketing a mixture of proprietary and open source products which gives rise to new challenges for which there are no immediate solutions. The SCO litigation involves unresolved allegations revolving around the alleged misappropriation of SCO’s Unix technology. Whilst the litigation seems to have led to an increased awareness and popularity of Linux systems, at a practical level it has caused Centrelink to refrain from deploying the open source software until the matter is settled.