The patent eligibility of software and business method patents often presents a difficult problem: how do we allow inventors to protect computer-based innovations without pre-empting the public’s use of abstract ideas and algorithms that underlie these inventions? This question has vexed both patent offices and the courts in many jurisdictions in recent years. Recent developments in Canada and the United States suggest that we are not likely to see a bright line test in the near future.

Many observers had hoped that the U.S. Supreme Court’s 2010 decision in Bilski v. Kappos would provide definitive guidance, at least in the United States, for determining when software or business methods are patentable. However, the Court in that instance declined to elucidate any new tests, or even to overturn the “machine-or-transformation” test previously developed in the Court of Appeals for the Federal Circuit. Rather, the Court left the door open for future wrangling by calling the machine-or-transformation test a “useful and important clue” but not the “sole test”.

Last year, the U.S. Supreme Court again found itself considering patentable subject matter in Mayo v. Prometheus. Although Prometheus dealt with pharmaceutical patents, the Court’s reasons were nevertheless applicable to computer-based inventions. In particular, the Court held that patent claims must not pre-empt or monopolize laws of nature, and that any elements of a claim that seek to narrow the scope of a claim (so that it no longer broadly covers a law of nature) must be more than routine or conventional elements in the relevant field. Some commentators expressed concern at this result, noting that it might cause uncertainty in the software field, where claims can sometimes be reduced to algorithms implemented by a computer. If the algorithm is considered to express a “law of nature” or “naturally occurring relationship”, and if implementation by a computer is considered a routine step, then many software claims could be found ineligible.

Following the Prometheus decision, the Court of Appeals for the Federal Circuit (CAFC) – the senior patent court in the United States – revisited its earlier decision in CLS Bank v. Alice Corp. In an en banc decision issued in May 2013, the Court found that claims relating to a computerized trading platform were not patent eligible, reasoning that the claims were essentially directed to a form of escrow – an abstract idea – and that using a computer to perform the recited operations failed to provide an “inventive concept” sufficient to merit patentability. However, the individual judges’ reasons varied widely, as the ten-judge panel issued five different sets of reasons in a 135-page judgment that was widely interpreted as inviting the Supreme Court to take on the case.

While the Prometheus decision was cause for concern in the computer industry, the CAFC decision in CLS Bank truly set off alarm bells, suggesting that we may be entering a new era in which the courts and the United States Patent Office (USPTO) would curtail the availability and scope of software and business method patents.

However, only one month later, the CAFC issued another decision suggesting that software and business method patents are not quite dead yet. In Ultramercial v. Hulu, the CAFC acknowledged that “[d]efining ‘abstractness’ has presented difficult problems” and clarified that “a claim is not patent eligible only if, instead of claiming an application of an abstract idea, the claim is instead to the abstract idea itself” (emphasis in original).

To distinguish between applications of an idea, and the idea itself, the CAFC suggested that “the relevant inquiry is whether a claim, as a whole, includes meaningful limitations restricting it to an application, rather than merely an abstract idea.” A claim is not meaningfully limited if it covers all practical applications of an abstract idea, or contains only insignificant post-solution activity. But “meaningful limitations may include the computer being part of the solution, being integral to the performance of the method, or containing an improvement in computer technology.”

In the specific case of Ultramercial, the CAFC ruled that patent claims to a method of advertising on the Internet were patent eligible, at least because they “plainly require that the method be performed through computers, on the Internet, and in a cyber-market environment.” The claims did not, according to the Court, pre-empt all uses of the abstract idea underlying the invention.

Casual observers can be forgiven for wondering how to draw the distinction between an unpatentable computerized trading platform and a patent-eligible online advertising system. As U.S. courts have demonstrated, this issue is likely to remain a grey area for the foreseeable future. Nevertheless, the courts have provided several clues as to which claims that are likely to remain patent eligible regardless of where the line is eventually drawn. First, patent claims for computer-based inventions should be crafted to ensure that they do not merely describe an abstract idea or mathematical formula, but rather specific applications of the idea. Where possible, the meaningful, integral use of a computer processor or other “machines” should be explicitly recited in claims. Finally, care should be taken to ensure that the claim does not pre-empt all practical uses of the underlying abstract idea. Prospective patentees would do well to consult with a skilled patent practitioner as, ultimately, the specific language to be used will depend greatly on the nature of each invention.