This week, Sen. Chuck Schumer, D-N.Y., announced plans to introduce a bill that would mandate the incorporation of geo-fencing technologies into all domestically produced commercial drones, which would prevent them from operating within two miles of any airport and above 500 feet. Schumer believes this proposal offers an “elegant solution” to the recent rise in drone “incidents.” The reality is that the senator is fishing for the moon in the water. The same impulse toward preemptive regulation applies to autonomous vehicles. But there aren’t yet any real problems with drones or autonomous vehicles that call for regulation. These technologies are babies still in the cradle. Legislators need to be careful not to strangle them with red tape.
There have been 700 documented incidents of drones flying near piloted aircraft so far this year. At first glance, that might seem like a public safety nightmare. However, of those 700 “incidents” only 25 involved “near-collisions” with aircraft, and no actual collisions have been reported. To put it into perspective, nearly 14,000 animals actually collided with airplanes in 2014. Before turning to the drone “problem,” Sen. Schumer ought to address the danger posed to aircraft by wildlife.
In any case, the senator’s proposed legislation is likely to be either redundant or counterproductive. NASA and private firms are already working together in an effort to create an automated air traffic control system for drones. Federal legislation at this early stage could unintentionally retard efforts to integrate commercial drones into the national airspace.
Geofencing technology, which uses GPS tracking technology to prevent drones and other airborne platforms from entering a predefined geographical area, may or may not be the best way to keep the public safe or sensitive facilities secure. It’s possible that new technologies will come along and prove themselves better suited to the job. Mandating geofencing technology today may close off the development of better solutions later.
Schumer’s legislation also fails to adequately guard against terrorists or criminals who may try to use drones to damage airports and other sensitive sites. Kevin Poulsen, writing in Wired, notes that, “Nobody willing to strap a bomb to a toy drone will be deterred” by a geofencing mandate. Schumer’s proposal is lazy. It attempts to police drone use, yet fails to address the tough problems drones pose for public safety and security, such as how to effectively curtail nefarious users from using them near sensitive locations. Its main effect would be to interfere with the ongoing efforts of public-private partnerships to deal with the security and safety issues of drones while upsetting drone hobbyists who live near airports.
Premature regulation also threatens to hamper the development of self-driving vehicles. Earlier this year, the office of Sen. Ed Markey, D-Mass., released a report that assessed the current state of privacy and security in the autonomous vehicles industry. The report is the primary basis for the SPY Car Act of 2015, which Markey has sponsored with Sen. Richard Blumenthal, D-Conn. The report, and therefore the legislation, is based on flawed assumptions, which I identified in a blog post earlier this year.
The report cites only two specific instances of car “hacking.” Neither scenario involved “hacking” in a nefarious sense, or justifies a call for federal regulation. In the first case, a private firm voluntarily removed an app from the Google Play store as a “precautionary measure,” despite an analysis that failed to “indicate any ability to introduce malicious code or steal data.” The second case may technically count as “hacking,” but isn’t very worrying. Car owners have been attempting to reprogram the onboard computers of their own vehicles. It is illegal for car owners to hack their steel carriages, but many decry these rules as unfair and anathema to hobbyist tinkering and innovation more broadly.
Federal regulation isn’t necessary to address the mere possibility of security vulnerabilities. Auto manufacturers and technology firms have every reason to address issues of security and privacy in the designs of their autonomous vehicles before problems arise. Intel has announced the creation of the Automotive Security Review Board “to research and collaborate on ways to improve automotive security products and technology.” The company’s own white paper on the challenges posed by an increasingly interconnected roadway clearly shows that, far from being ignorant of the problems before them, companies are proactively building privacy and security into these new vehicles by design.
Markey’s legislation attempts to regulate connected car technologies in an ex ante fashion, approaching the “problems” of privacy and security as ones that require regulation before any harm has come from current market practices. Adam Thierer of the Mercatus Center refers to this approach as precautionary principle regulation. It makes more sense to pass regulations that deal with problems as they arise, and only then if legal remedies prove incapable of handling market failures. Regulating new technologies over fears of what might happen will stifle innovation. We ought to reject the precautionary principle in favor of permissionless innovation, the idea that innovators and entrepreneurs need not ask or get permission before trying new things.
New technology is often susceptible to “panic cycles.” The public can be taken in by “fear entrepreneurs” seeking to advance their own agendas. However, once the benefits of emerging technologies become clear, public attitudes tend to adjust to accommodate the change innovation has brought about. (For more on panic cycles, see this new paper from the Information Technology and Innovation Foundation and this Cato policy brief from Larry Downes.) Regulators and lawmakers would do well not to act out of fear. They should recognize that they cannot possibly have all the answers, or even especially good answers, to the questions raised by unfamiliar, rapidly-developing new technologies.