Regulating Drones Is as Bad an Idea as Regulating 3D Printing
Last week, my colleagues Adam Thierer, Eli Dourado, and I filed comments in a Federal Aviation Administration proceeding urging the agency not to impose privacy regulations on commercial drones as they begin to be integrated into the national airspace. The general reaction from many of our friends in technology policy circles was essentially: “Are you crazy?!” That surprised us a bit since our argument was one of “permissionless innovation,” a pretty popular concept with that crowd.
As the Internet has demonstrated, innovation thrives when tinkerers, artists, and entrepreneurs can create without first having to ask anyone for permission. This is why, for example, so many of us oppose the excesses of copyright and patents — government regulations that create unnecessary and inefficient permissions-based barriers to entry.
Some argue that we can have both regulation and innovation; that the two things are not mutually exclusive. That’s certainly true, but regulation does involve a tradeoff, even if only on the margin. By regulating you won’t get the same amount or kind of innovation than you would without regulating, and while the choice to regulate may well be justified, one can’t pretend the tradeoff doesn’t exist.
Regulation, by definition, involves making certain uses of a technology verboten, or at least more costly because an innovator must first get the regulator’s permission. By regulating, there are some innovative uses of the regulated technology that you will necessarily be foreclosing. Again, this choice might be justifiable, perhaps in defense of values such as privacy. The trick, though, is making sure that you regulate in a way that gives up no more innovative potential than is necessary to achieve your goal.
But here’s the rub: In the case of new technologies, and especially new technologies that are platforms for innovation, such as the Internet and drones, we cannot know ahead of time what innovations we will be foreclosing by regulating. Let me say that again: we cannot know, so we cannot possibly conduct an informed cost-benefit analysis to determine if the tradeoff is worth making. Any attempt at such an analysis is necessarily based on hypotheticals about the harms that will be addressed by regulation and the hypotheticals about what will be given up.
It is better to eschew such a precautionary approach and wait to see what happens. Allow the technology to develop. Allow entrepreneurs and creators to experiment and see if any harms do indeed materialize. See if existing laws, norms, and other structures can handle any real harms that develop. And then only if regulation is shown to be necessary does one regulate, relying on more information about the real tradeoff to be made, and safer in the knowledge that one is not foregoing more innovation than necessary.
Drones are different. That is the argument that our friends at the ACLU, CDT, EFF, and EPIC made arguing for strict privacy provisions in their comments to the same FAA proceeding. Because drones are substantially cheaper than manned aircraft, because they can stay aloft for much longer, and because they can go places previously unreachable by manned aircraft, goes the argument, the technology is radically different in kind from any prior technology. As Amie Stepanovich has put it, smartphone cameras are to drones as binoculars are to an observatory’s telescope. As a result, they argue, this new technology poses privacy risks unlike any we have known before — on a massively large scale — and thus a precautionary approach is in order.
What we need in order to assess the need for such a regulatory response is a general framework to analyze risky new technologies. Luckily, Michael Weinberg of Public Knowledge has developed just that framework in his pioneering work thinking about 3D printers.
Like drones, 3D printers are not just a new technology, but a new platform for innovation and personal empowerment that also has potentially nefarious uses. In the case of 3D printing, some have suggested regulating the technology because it can be used to produce guns.
“The first question that we should ask is ‘is this actually new?’,” writes Weinberg. ”In other words, is 3D printing allowing people to do something that they were unable to do before, or is this simply getting attention because someone got around to doing it with a 3D printer?”
He says the answer is no. People today already trade digital designs for gun parts and make them at home using CNC lathes and other milling machines. The same can be said about the type of surveillance or invasion of privacy some fear drones will facilitate. Today you can already engage in spying and surveillance without a drone.
“[P]resumably our national firearms policy already recognizes that people can make weapons at home and has structured rules accordingly,” Weinberg continues. And as Adam, Eli, and I point out in our filing, there is an existing structure of privacy torts and Peeping Tom laws that already prohibit the kinds of harmful activities that for which drones might be used.
Weinberg continues in his framework: “The second question is ‘even if it is not new, does it fundamentally change the existing activity?’” He again again answers in the negative. “People were downloading files used to make firearms on automated (or semi-automated) machines before they had access to 3D printers and, at this stage, it is hard to see how this changes that activity.” The same is true of drones. People have been spying on others before they had access to drones, and potentially doing so not just on foot or from a car, but using a radio-controlled aircraft. The introduction of drones doesn’t change this fact.
And finally: “The third question is ’is it possible to fundamentally change the existing activity in the future?’” This is the most important question, and the one that the privacy advocates seem to be getting at. Weinberg says the answer for 3D printer is ”maybe,” and I’d say it’s the same answer for drones.
Today there are probably more CNC milling machines available to the public than 3D printers, Weinberg says, but you can imagine a future in which 3D printing is incredibly ubiquitous and accessible, and in which it would be so easy to download and print a gun that new policies regulating 3D printers might be justified. The same is true of drones. We can certainly imagine scenarios that would justify regulation in the name of privacy.
Yet Weinberg understands why a precautionary approach can be so pernicious:
But that is the real challenge with all policy connected to 3D printing (and to emerging technologies more generally): being able to imagine a way where a technology could be misused in the future is not a sound basis for policy, and certainly not a sound basis to limit its growth.
Imagining a futuristic 3D printing dystopia and then trying to create policies to stop it will inevitably be a counterproductive exercise for at least two reasons. First, the imagined dystopic future will never actually happen. If we were good at predicting how new technologies would impact society we would all be rich. That makes new legislation designed to prevent the bad future a waste of time at best.
Second, and more problematic, is that any legislation aimed at preventing an imagined future is much more likely to block unexpected positive developments. We do not know how 3D printing will actually impact society, but we can be fairly sure that today’s projections will seem laughable 10 or 20 years from now. Laws enacted during a time of 3D printing anxiety are much more likely to prevent good things than block bad things.
Substitute drone for 3D printing, and I could not have said it better myself. Regulation in the name of safety, to protect privacy, or to protect intellectual property rights will always involve a tradeoff. It means giving up, at least partially, what might otherwise be a permissionless space for innovation. That may be necessary sometimes, but we should resist doing so preemptively — even when we think the new technology is unlike anything we’ve seen before, and even when we can dream up dreadful possible scenarios that it might facilitate — because we can’t possibly predict what we would be giving up.
On both 3D printing and drones, let’s not regulate until we’re fairly confident we have no other choice. Otherwise, despite our best intentions, we might be leaving our children a worse possible world.
