In a nutshell, the question is: how are the rights of a patent owner weighed up against those of other people or businesses, which may have been using similar technology for some time prior to the filing of the patent application, but which are then subsequently accused of infringing the patent? And how does this relate to the concept of ‘prior art’?
James was specifically interested in the case of a software-based invention, however the Australian patent law makes no distinction between different fields of technology. Nonetheless, it is probably fair to say that software inventions can, in some cases at least, be more easily ‘used’ without it necessarily being apparent how the software works, or what exactly it does, which has some impact on its prior art effect, so we will address this specific issue towards the end of the article.
The short answer is that under the Australian Patents Act 1990 a prior use of an invention might be used in one, or both, of two different ways to fend off an accusation of patent infringement. Firstly, it might be possible to employ the prior use as a form of prior art in order to attack the validity of the patent claims. Alternatively, section 119 of the Act provides that, in appropriate circumstances, the use of a patented invention before the priority date of the patent claims acts as a defence to an accusation of infringement.
There is also a ‘shortcut’ tactic, known as the ‘Gillette Defence’, which essentially says that if the accused infringer is doing nothing more than what was already done before the priority date of the patent claims, then either the patent is invalid, or the claims are not infringed. The defence runs that it does not matter which, because either way the accused is off the hook! This is a shortcut, because in principle it enables the court to dismiss a patent case without actually addressing the substantive questions of infringement or validity. There is, however, considerable doubt as to whether this is a legitimate form of legal reasoning under the current Act – quite aside from anything else, it has the potential to leave an invalid patent on the Register, to be used again another day! This issue arose in the recent Australian case of Otsuka Pharmaceutical Co., Ltd v Generic Health Pty Ltd  FCA 239, which we wrote about in Australian Federal Court Blocks Generic ABILIFY (Aripiprazole), so we will not dwell on it again here.
PRIOR USE AS PRIOR ARTAn invention claimed in a patent or application is assessed for novelty and inventive step against the relevant ‘prior art base’, which is the body of information that was made available to the public (or at least to some of the public) before the applicant’s earliest established filing date (i.e. the priority date).
Most commonly, particularly in the normal course of patent examination, the prior art relied upon is in documentary form, i.e. it consists of one or more publications (including electronic documents) which can be shown to have been published before the priority date.
However, the prior art base in Australia also includes ‘information made publicly available through doing an act’ (see definition of ‘prior art base’ in Schedule 1 of the Patents Act 1990) anywhere in the world. The ‘doing [of] an act’ includes any public use of a relevant product or process.
In Jupiters Ltd v Neurizon Pty Ltd  FCAFC 90, the Full Court of the Federal Court of Australia (on appeal) had this to say about the requirements for establishing invalidity on the basis of an alleged prior use occurring independently of the patentee (at ):
What is required to satisfy the provision is an ‘enabling disclosure’ of the invention by the prior conduct…
This means that the act must reveal enough detail for a non-inventive skilled worker in the relevant field of technology to put the invention into effect. It may therefore not be sufficient that something is used in public before the priority date, it that use in not revealing enough.
Finding that there was no recent authority on this issue under the Australian Patents Act 1990, the court in Jupiters turned (with appropriate caution) to some UK decisions for guidance, making the following observations about those cases:
- the demonstration of a computer graphics system in operation, without disclosure of the program code, or any detailed explanation of how it worked, was not an ‘enabling disclosure’;
- disclosure occurs if the information is made available to at least one member of the public who is free in law and equity to use it, i.e. not subject to an obligation of confidentiality;
- where a product is made available without relevant restrictions, a skilled person may be assumed to be capable of ‘reverse engineering’ the product to determine how it works, so long as this does not involve undue burden; and
- the public use of a prototype traffic light controller constituted a relevant prior use, even though the controller housing remained locked, because its operation could be inferred by the skilled person simply by observation over a sufficient period of time.
The Jupiters case is thus a good demonstration of the difficulty that can arise in a prior use attack on a patent. In order to invalidate a patent, it is not enough to show that a relevant prior use occurred. It must be established by admissible evidence, and with sufficient confidence (i.e. that it is more likely the case than not), that the prior use resulted in an enabling disclosure of the invention. In the absence of a specific prior art publication, much of the evidence relevant to prior use will necessarily come from witnesses to the use, and/or from technical experts. Jupiters was unable to meet this evidentiary burden, even with regard to its own prior use!
THE PRIOR USE DEFENCEAs the Jupiters court also pointed out (at ), the general principle of a ‘prior use’ defence, historically-speaking, was that ‘a third party cannot be prevented from continuing to do what they were doing prior to the grant of the patent’, which seems like a perfectly fair and reasonable proposition.
There is, however, an alternative theory, which can be derived from one of the policy justifications for having patent laws in the first place: society benefits when new technology is disseminated; this benefit is reduced if inventions are kept secret; patents therefore provide an incentive to disclose (i.e. in the patent specification), in exchange for a limited monopoly; therefore those who choose to conceal an invention from the public, by electing not to disclose and/or to apply for their own patent, should perhaps not have the benefit of a defence.
The complete absence of a prior use defence is no doubt an extreme position, however it has long been the basic rule under US patent law, for example (although the recent America Invents reforms introduce an expanded prior use defence, along with the change to first-inventor-to-file).
The position under section 119 of the Australian Patents Act 1990 represents a compromise. Generally speaking, prior use is a defence, subject to some limitations. Specifically, the defence will normally be available if the accused infringer was, immediately prior to the priority date of the claim(s):
- exploiting the claimed product, method or process in Australia; or
- had taken definite steps (contractually or otherwise) to exploit the product, method or process.
- stopped (except temporarily) exploiting the claimed product, method or process in Australia; or
- abandoned (except temporarily) the steps to exploit the product, method or process.
The prior use defence is ‘personal’ in nature, in the sense that it only applies to an accused infringer which was itself using the invention before the priority date. A third party cannot rely on the prior use of another as a defence, unless it is an enabling public use which can be used to invalidate the patent. Interestingly, however, under subsection 119(4), the prior use defence is transferrable to another party, for example in the case that the business is sold wholly or in relevant part to a successor.
IS SOFTWARE A SPECIAL CASE?As noted above, the law draws no distinction between different fields of technology. However, as a practical matter processes in general are more likely to be the basis of a prior use defence, or a contested prior public use attack, simply because it is easier to use a process without disclosing its details. Most products, once sold and available for inspection by the public at leisure, are susceptible, at least in principle, to reverse engineering.
For example, it is entirely feasible, with careful control of access to equipment and documentation, to maintain a new an improved process for manufacturing a chemical or other product as a trade secret.
Digital systems, including those implemented principally in software, are generally best characterised as processes rather than products. They implement particular functionality – such as graphics processing, controlling traffic lights, or operating gaming machines (to use the examples from the cases discussed above) – and the specific hardware and software components used to implement that functionality is generally interchangeable. It is, for example, irrelevant whether the locked controller box in the traffic light example contains discrete digital logic components, a programmable logic device, or a microprocessor with appropriate software programs. The functionality of the traffic lights is just the same.
Software, in this sense, is a ‘black box’. It is usually possible to see what it does, but not always apparent precisely how it does what it does. And while digital systems are theoretically susceptible to reverse engineering, it is technically possible to make this very difficult – particularly if the entire hardware/software system is integrated into a single chip.
So while software is legally no different from any other technology, it might well be more likely that a prior use defence would arise in relation to a software, or other digital, system. This is, of course, where open source software provides a potentially rich vein of largely untapped prior art, but that is a whole other topic, perhaps for another day…
CONCLUSIONPrior use, if it is public, is potential prior art which can be used to attack the validity of a subsequent patent or application. However, if the use has not resulted in an ‘enabling disclosure’ – or, more to the point, this cannot be proven to be the case to the standard required by the courts – then a patent may validly be held by a independent inventor, who happened to make the decision to file an application.
In some cases, maintaining a trade secret may be a more profitable course of action that filing for a patent. In theory, something such as the secret recipe for a cola drink, or for the coating applied to fried chicken, may be uniquely maintained for far longer than the 20 year maximum term of a patent! And if you want a patent, you must disclose fully, so that your invention is free for the public to use once the patent has expired.
But trade secrecy has its risks, not the least of these being that someone else may obtain a valid patent. After all, it is your choice not to disclose your invention so that it joins the prior art base, or to protect it with your own patent. In Australia, however, you can take some comfort from the fact that your own prior use will provide you with an infringement defence, even if it cannot be used to invalidate a patent.