For the first time in more than 30 years, the US Supreme Court is tackling one of the most vexing questions in patent law: can software be patented? Its decision could affect the future of personalized medicine.

On 31 March, the court will hear arguments in Alice Corporation v. CLS Bank International, a lawsuit over four patents on software intended to reduce risk in financial transactions. The case is the latest example of Supreme Court efforts to rein in the US Patent and Trademark Office, which is notoriously permissive in granting commercial protection to inventions. Two years ago, the court struck down a class of patent often used to protect medical diagnostic tests. Last year, it ended a 30-year tradition of issuing patents on naturally occurring genes.

With the latest case, the court has the potential to restrict software patents, with repercussions not only for the technology industry, but also, yet again, for medical-diagnostics companies, says Jonathan Masur, deputy dean of the University of Chicago Law School in Illinois, who specializes in intellectual property. “It’s enormously important,” he says.

Software patents are controversial. Critics charge that they tend to be vague and cover obvious inventions, and that they hinder innovation. They are magnets for ‘patent trolls’ — businesses that collect broad patents mainly to force others to pay licensing fees. Some critics also argue that computer programs are simply the expression of mathematical formulae, and thus can be considered abstract ideas, which, along with laws of nature, are not patentable.

Yet in 1981, the Supreme Court conceded that some software could be patented, and since then the number of software patents has soared.

The Alice case began in 2007, when CLS Bank International, based in New York, sued Alice Corporation of Melbourne, Australia, charging that four of Alice’s patents were invalid. Alice countersued CLS Bank for infringing the same patents. But CLS Bank could be in a good position, if the Supreme Court’s interest in limiting patentability in recent years is any indication (see ‘Valid concerns’). Observers are hoping that the court will use the case to develop concrete criteria to determine which software patents are valid. “The problem with the law to date is that it’s very subjective,” says Robert Sachs, a partner at the law firm Fenwick and West in San Francisco, California. “Every judge gets to decide: ‘Is this the kind of thing that I think a patent claim ought to cover?’”

To bolster its arguments, CLS Bank is citing a 2012 case, Mayo Collaborative Services v. Prometheus Laboratories, that struck down two medical-diagnostics patents and sent tremors through the diagnostics industry. The patents covered the act of measuring metabolites to determine the proper dosage of certain drugs used to treat autoimmune diseases. The nine judges on the Supreme Court unanimously decided that the patents merely laid claim to a law of nature — the body’s breakdown of the drugs — and were therefore invalid.

The Mayo ruling represented a fundamental change, says Brian Dorn, a patent attorney at the Barnes & Thornburg law firm in Minneapolis, Minnesota, who counts diagnostics companies among his clients. “It’s been hard to get around,” he says. “We’ve been getting patent rejections based on that case.” Some even blame the court’s decision for a drop in US venture-capital investment in medical-diagnostics companies — from US$395 million in 2011 to$278 million in 2013, according to London-based consultancy firm Pricewaterhouse­Coopers.

Mayo is also setting a precedent for lower courts. Last October it was used, along with the 2013 Association for Molecular Pathology v. Myriad Genetics ruling that invalidated patents on genes (see Nature 498, 281–282; 2013), to throw out key patents on a prominent non-invasive prenatal test for Down’s syndrome, held by the diagnostics company Sequenom of San Diego, California. The ruling, by a California district court, used a strict interpretation of Mayo, says Christopher Holman, a law professor at the University of Missouri–Kansas City Law School. “If that is how the courts are going to use that case, it’s a very bad sign for industry,” he says.

The Alice case could be a chance for the Supreme Court to clarify how Mayo should be interpreted, says Masur: rulings on abstract ideas (as is likely to arise from Alice) are often applied to arguments about patenting laws of nature (as in Mayo). If the Supreme Court invalidates many kinds of software patents, for example, diagnostics firms could be in for more trouble, he says.

The Supreme Court’s tendency towards restricting patents is a concern for Gregory Graff, an agricultural economist who studies intellectual property at Colorado State University in Fort Collins. He argues that the US patent system should not be reformed through a series of radical court decisions. Instead, he would prefer to see legislation and patent-office procedures tweaked as fields of technology evolve. “You don’t need the nuclear option of blowing a hole in what’s deemed patentable,” he says.