Earlier this year, a group of 51 legal scholars and economists sent a letter to the US Congress urging it to take action on the increasing toll of frivolous patent lawsuits. Over the past five years, they said, researchers have published more than two dozen studies on the economic consequences of patent litigation. The view that has emerged is grim: the lawsuits are hindering research and development, and slowing the launch of firms.

Less than a month later, another 40 scholars rebuffed the claims, saying that the impact of the lawsuits has been exaggerated. Furthermore, they argued, patent litigation is on the wane, and legislation to rein it in could damage the US “engine of innovation” by weakening patent protections for inventors.

Such are the muddied waters that Congress has been navigating as it seeks to respond to the cries of technology companies and of President Barack Obama’s administration, which want to crack down on lawsuits launched by ‘patent trolls’. No fairy tale, these entities are essentially holding firms to ransom, threatening organizations that are making use of the innovations with expensive, time-consuming lawsuits if they do not pay to license the patent. A 2013 attempt to curb such legislation met with failure last year. Lawmakers now seem to be making progress (see page 270).

Much of the scholarly debate boils down to a difficulty that has also plagued Congress: how to define a troll. Universities, too, license their patents, often for a fee, to those who want to use their researchers’ inventions to create a product or service. As such, they are considered ‘non-practising entities’, a more-polite term than troll, but the two labels are often used interchangeably.

Scholars generally argue that universities should be considered differently because they work towards a social good and their patenting efforts spur innovation based on academic discoveries. This is in stark contrast to a troll, which accumulates weak, broad patents with the sole intent of using them to push firms into settling a lawsuit before the expense of the litigation damages their business. Lawmakers in the US Senate seem to agree with this distinction, and last month created a carve-out that excludes universities from some of the proposed measures for cracking down on patent trolls.

It is important not to see patent-troll legislation as a panacea.

But the distinction has fuzzy boundaries: some universities are highly aggressive in monetizing their patents, even licensing them to companies that are considered to be trolls (see Nature 501, 471–472; 2013). Earlier this year, the Association of American Universities and the Association of Public and Land-grant Universities took a step in the right direction by urging their members not to align with trolls. Universities should heed that guidance or risk losing the faith of Congress and the public. The Senate loophole for institutions of higher education was a political necessity in the face of heavy lobbying by universities, but that lobbying would have been much less persuasive had it not been tied to widespread public trust.

As Congress has wrestled with definitions, its overall approach for deterring frivolous lawsuits has remained fairly constant: make them more risky for the plaintiff. It is a welcome change to a system that is much too easy to exploit, but it is a blunt tool that could jeopardize the ability of small firms to defend their intellectual property. And even if it succeeds in Congress, it will not tackle the underlying problem: the US Patent and Trademark Office is granting far too many vague and redundant patents. This is a particular problem for software, but affects other fields, too.

Measures to raise the bar — including a process that allows parties to challenge a patent without needing to resort to litigation — may be having an effect: the number of patent lawsuits dropped by 18% between 2013 and 2014. But it is important not to see patent-troll legislation as a panacea. Fundamental changes at the patent office remain the key to curbing abuse.