Smart Meters On Your Home: Where Does The Data Go?


Patricia Burke of local activist group Halt MA Smart Meters brings to our attention an effort by the Massachusetts Department of Public Utilities to study how to achieve universal adoption of smart meters on residential homes in Massachusetts. This implicates the Fourth Amendment because electricity usage within a home, if gathered many times over a 24-hour period and transmitted to a government agency, may constitute a warrantless search of that home to which the home’s resident has not consented. The IT Law Wiki provides an excellent overview of the constitutional issues here.

It is unclear from the documents provided by DPU whether any plan actually exists for what to do with the data gathered by utility companies such as National Grid. It is possible that the data would remain with the utilities, and would be used to implement peak pricing that in turn would both increase profits and reduce stress on the grid at peak times. However, as we have seen with the NSA scandals, it is very easy for government agencies to get court orders requesting the ongoing release of such records by utilities to law enforcement. The records would, under the “third-party doctrine,” probably be considered to be the property of the utility company, and therefore law enforcement would typically not seek either the permission of the resident or a duly executed warrant from a judge before accessing this data.

Long-time readers of this blog will know where we’re going with this. Yes, it’s fusion centers.

This type of data would be a classic example of the kind of large database that, without explicit prohibitions, can find its way to the Commonwealth Fusion Center or Boston Regional Intelligence Center, where it can be integrated with data from many other sources and used to implicate defendants. Yes, utility data could also be used to exonerate defendants, but let’s get real here: the people with the data are on the side of the prosecution, not the defense, and it would take a savvy defense lawyer, like this one in Florida, to get a judge to require production of utility data in order to prove that their client was or was not at home at the time of the alleged offense.

The most relevant Supreme Court precedent here is Kyllo v. United States, 533 U. S. 27 (2001), which held that “Where, as here, the Government uses a device that is not in general public use, to explore details of a private home that would previously have been unknowable without physical intrusion, the surveillance is a Fourth Amendment “search,” and is presumptively unreasonable without a warrant.” Litigation brought on Fourth Amendment grounds would therefore presumably have to show (a) that smart meter data had been shared with law enforcement for that home, without the knowledge or consent of the resident, and that (b) smart meters are a “device that is not in general public use.” The difficulty with the “general public use” test, which was new in Kyllo, is that the Justices did not provide further definition of what that test meant. One could easily argue that to be in “general public use”, a smart meter had to be generally available for sale to members of the general public. A smart meter add-on to a regular meter is available on Amazon, but the kinds of meter likely to be litigated would be meters installed by the utility company themselves, not a discretionary add-on device purchased by a homeowner. But it’s fair to say that courts could apply this very general language either way when it comes to smart meters.

One deeper problem here is that the “general public use” test is essentially a nonsense. A technology can move from early-stage production to wide adoption by the general public inside of a year. To take one emerging technology, there’s no particular reason outside of Kyllo to think that, say, a search conducted using a gigapixel camera would require a warrant now but would not require a warrant at the end of this year if their use had exploded by then. It would have been better for the Fourth Amendment and clearer for law enforcement if Kyllo had been decided the same way, using a less arbitrary and time-bound test.

Currently, the DPU’s discussion document does not mention the Fourth Amendment, and does not explicitly consider Fourth Amendment issues to be one of the key questions to be resolved. But what happens to this privacy-implicating data does matter. The simplest and oldest rule here is also be the best: if a private company is going to share private information about you with the government, it should only do so if the government produces an individual, probable-cause-based warrant first.

The DPU does indicate that they will be holding a future “proceeding” to investigate privacy and cybersecurity concerns:

The Department established its current practices concerning the protection and release of electricity usage data to promote the competitive electricity market and to ensure customers’ privacy and allow customers to request the collection of their interval data. Competitive Market Initiatives, D.T.E. 01-54 A, at 9-12 (2001); Installation of Advanced Metering Equipment for Residential Customers, D.T.E. 01-28, at 8 (2001). However, the future collection of interval data from all customers through advanced metering functionality raises additional potential privacy concerns. While protecting customer data must be a high priority for electric distribution companies, electricity usage and consumption data must be available to customers, as well as to competitive suppliers and other service providers if authorized by customers, in order to fully realize the benefits of a modern grid. The Department must investigate these issues and strike an appropriate balance between ensuring safeguards and enabling customers to realize the full potential of interval data and data management. Thus, in a separate proceeding we will explore how the Department should ensure: (1) protection of customer data privacy; (2) access to data by customers and authorized third parties; (3) timing and availability of data; and (4) uses of aggregated interval data.

We look forward to it.

Leave a Reply