Sensors and Journalism

By Nabiha Syed When Samuel Warren and Louis Brandeis first wrung their hands over “Kodakers lying in wait,” they likely didn’t realize that the camera was just the first of many sensors that would come to document the human experience. Sensors allow journalists, as Matthew Schroyer explains, to “arrive at evidence-based conclusions in the course of a journalism investigation.” Both the “evidence-based conclusions” drawn and the process of “investigation” yield separate legal issues. This piece will address potential issues related to consent, which is an issue at the investigation stage, as well as defamation and false light, both of which relate to the publication of sensorsourced content. Defining Sensors At its most simple, a sensor is a device that detects and responds to some type of input—light, heat, motion, moisture, electronic signals, pressure— from the physical environment. Under this definition, cameras are sensors: A camera detects and captures light input to capture an image. But while courts have grappled with the sensor-as-camera for over a century now, especially in the newsgathering context, other types of sensors have not received the same attention.

It is an open question whether courts will approach sensors as a logical outgrowth of cameras, or will treat them as a new category entirely. On one hand, sensors can capture what happens in public, just as cameras can; some sensors are just capable of responding to input other than light. On the other, however, sensors pick up information that humans cannot discern— which is unlike the visible image, and the relationship between the camera and the public. Courts could very well construe this difference to mean that some sensor use is more invasive than cameras for image-capturing. Cameras are much better understood than sensors by the general public, leading to some issues with whether sensor use would violate some reasonable sense of privacy. Newsgathering with Sensors: Consent and Notification Issues American courts have long rejected liability for observation that takes place in public thoroughfares or in plain view, where one could be observed by passersby.2courts’ reasoning on sensor use: So long as the sensors are in public, picking up information readily available in the public thoroughfare, the usage is legal. By this logic, explicit consent for sensor use is unnecessary, because the input detected by the sensor is public. The recent “right to record” cases underscore the value in recording what transpires in public. Several courts have now held that the First Amendment protects the right to record audio and video in public places, especially when recording the conduct of government officials.3record does not mean that consent is no longer necessary for recording. A closer look indicates that existing caselaw focuses on public officials, and even then draws the line where there is a “compelling interest,” such as in border security.4right does high-

light, however, is that courts recognize a fundamental interest in observing the public sphere, and are willing to extend protection accordingly. This interest may be helpful when thinking about using sensors to record particular types of input. Despite this trend, there are still two reasons that gaining consent or giving advance notification of the use of sensors is useful. The first is that courts could conclude that the unique capabilities of various sensors distinguish them entirely from cameras. After all, a camera senses and captures light to recreate, in fixed form, images that a person already has the capacity to see. That is, the sensor duplicates what a passerby can witness. This is not exactly true of many sensors, such as infrared or pressure tools. Courts may then refuse to extend the right to record to sensors as a blanket rule. The second is that any right to record or gather data—image or otherwise— is still subject to the longstanding principle that journalists are bound by laws of general applicability. As the Supreme Court has explained in Cohen v. Cowles Media Company, 501 U.S. 663 (1991), First Amendment rights are not offended when “generally applicable” laws have “incidental effects” on the ability of the press to gather the news. In short, the press has no special exemption from the law. That means that both traditional tort law and other statutory protections apply to the newsgathering context. The following are two arenas of particular note to sensor journalism. Trespass Asking permission before one attaches or otherwise uses a sensor may ward off potential trespass claims, which could come about in two ways. Most obviously, one could affix a sensor to property owned by someone else: This scenario is precisely what the Supreme Court considered in United States v. Jones, where law enforcement attached a GPS sensing device to a suspect’s automobile and used the device to monitor the suspect’s movements. At

least some justices found that attaching the sensing device did, in fact, run afoul of common-law trespass.5signals one simple limiting factor for sensor journalism: Don’t attach your sensors to property you don’t own. Similarly, a sensor-equipped device could also fly or flow through private property, picking up a possible trespass claim along the way. A drone or a bicycle, as was used by the “Beijing Air Tracks” project to detect urban particulate and carbon monoxide levels, might have exactly this problem if operating on or through private property. Of course, permission to be on private property may be implied by custom or by the nature of the premises, such as a restaurant open during business hours. Implied consent may also be granted when one enters private premises and is not told to leave, or when prior contact indicates that one does not need express permission. This is subject to a reasonable person standard, and even this implied permission can always be revoked. Importantly, trespass protects only against physical invasion of property.6sensor travel through public property7data originating from private property, one would likely not face a trespass claim. And even if trespass-by-sensor occurs, a sensor journalist would only be responsible for the physical harm done—injury to reputation or emotional distress from the resulting publication often does not count. It’s hard to imagine what palpable damage to property the sensor itself can do. What is worth more attention is the vehicle, if any, to which a sensor is attached—a bicycle or a drone can cause more tangible damage.

Wiretapping Wiretapping conjures images of intercepting audio8or environmental pollution. But as the recent Google Street View Wi-Fi-sniffing case indicates, sensory collection that involves electronically transmitted data may run afoul of the Federal Wiretap Act. In the Google Street View case, the Ninth Circuit Court of Appeals determined that Google’s allegedly inadvertent collection of data—including usernames, passwords, emails, and documents—from an open (that is, non-password-protected) Wi-Fi network constituted a violation of the Wiretap Act. An exception in the Wiretap Act permits accessing electronic communications that are configured so they are “readily accessible to the general public,” which Google argued applied to unsecured Wi-Fi networks. Google also likened Wi-Fi signals to radio communication. The Ninth Circuit rejected these arguments, looking to what members of the general public might access: “[M]embers of the general public do not typically mistakenly intercept, store, and decode data transmitted by other devices on the network. Consequently, we conclude that Wi-Fi communications are sufficiently inaccessible that they do not constitute an electronic communication readily accessible to the general public.”

This logic could foreseeably extend to other types of sensory collection that gather electronic communications. Should a sensor be capable of such collection, it may be worth seeking consent of any possibly affected parties where the collection occurs. Printing Sensor-derived Content: Potential Defamation and False Light Issues At first glance, sensor-based reporting does not implicate reputational issues: If a sensor detects cicadas in a neighborhood, or measures the air pressure in a certain county, whose reputation is affected? And if the data is gathered in the aggregate, how would any conclusion attach to any single individual? Observations of natural phenomena that don’t implicate individuals or specific entities seem to skirt personality-oriented liability, and preclude claims like defamation and false light. The stories explaining sensory data might, however, be susceptible to these legal claims. Take the example of the drone flying over the Trinity River in Texas, whose camera picked up footage of a river full of pig blood downstream from a slaughterhouse. Imagine that—for follow-up purposes—you employ sensors in the water to detect the presence of animal blood or other contaminants. Your sensor detects particles of blood, and you conclude that based on past history, it is from the slaughterhouse. The slaughterhouse responds, arguing that there is another explanation for the conclusion you have drawn. You could easily face a defamation claim for communicating a false statement of fact that harms the slaughterhouse’s reputation. Defamation and false light claims are also important to understand because they will likely serve as proxy claims in jurisdictions, like New York, where common-law privacy claims are unavailable.

Defamation Truth is an absolute defense to any defamation claim, and a plaintiff carries the burden of demonstrating that a report is false. In any reporting that relies on sensor data, such as detecting blood particles in water, that data should be corroborated with further reporting to check the accuracy of any conclusions. One practice to avoid is any valorization of science: Sensorderived data should be verified and checked, just as a reporter would do with any tip from a source. As a general matter, a finding of defamation requires some degree of fault by the journalist. Under the First Amendment, the bare minimum standard of fault9prospective private libel plaintiff would have to show that a reporter’s conduct was less careful than one would expect of a reasonable journalist in similar circumstances. Cutting-edge forms of newsgathering complicate this analysis, because courts do not have many “similar circumstances” or existing precedent to compare. Presumably, courts will want to examine both whether the source (that is, the sensor) was reliable, and whether the source’s information (that is, the data) was reasonably interpreted by the reporter. This implicates both the reliability and accuracy of data collection as well as whether it actually substantiates an allegedly defamatory conclusion. How would a court engage in this analysis? Courts might look to whether a reporter has experience using sensors or in data analysis, whether a sensor was properly calibrated, or whether there was any reason to doubt its accuracy. The outcomes of this analysis could be wildly variable, depending on the court’s resources. Developing well-defined industry standards of practice will be crucial in helping courts determine whether there was any fault in the reporting process. False Light Unlike defamation, a false light claim—asserting that publicity cast a plaintiff in a false light in the public eye—requires no defamatory statements, only that a plaintiff finds the statements to be offensive. False light protects emotional injury and personal feelings like embarrassment or helplessness, regardless of whether any tangible harm actually occurred. Some states, including California, find that even untrue implications can be the foundation of a false light claim, even if no explicitly false statement has been made (as is necessary for a successful defamation claim). Flawed data, or flawed explanations of robust and accurate data, could lead to false implications. Imagine, for example, using a GPS sensor to trace and identify bus routes in a given city, like what Digital Matatus is doing in Nairobi. This data might yield details that some bus drivers are driving erratically or inefficiently. Should the article later focus on a particular bus driver, or include a photograph of one, that driver could easily object by claiming that the article implied that he too drove in such a negative manner, which cast him in a false light. Proving false light is a high burden: The alleged statement must identify the plaintiff, be highly offensive to a reasonable person, published widely, and demonstrate some fault on behalf of the publisher.10As discussed above, courts will likely face difficulty in parsing what fault looks like in the sensor context, precisely because there are few standards for comparison. To mitigate any risk, reporters should contemplate and closely analyze any implications that can be drawn from their work, including both the published data and its analysis.

Conclusions Sensors are not yet considered a routine newsgathering tool. This novelty may prove challenging to courts, which do not yet have the framework to assess the reliability or accuracy of sensors as a source, nor to evaluate whether a particular instance of sensor-based reporting was conducted reasonably. This is where ethics and industry practice must lead the way. Ethical guidance on how best to use, evaluate, and maintain sensors—and to interpret the resulting data—will not be found in the law. But ethical guidance and established industry practices can provide key instruction to courts, which will need assistance in determining whether a newsgatherer has relied on sensors in a responsible or reasonable way. In the meantime, the novelty and unique capabilities of sensors as a newsgathering tool may provoke legal claims.