any intrusion was unintentional or negligent. Similarly, an ‘interference with privacy’ requires the drone to have intruded to an unreasonable extent in the collection of personal information, and again relies on asymmetric information about whether information was even collected.
Faced with such uncertainties, a smaller proportion of cases will be pursued than would be the case if there were certainty about the filming, and some of the cases that are pursued will fail. As discussed earlier, the probability of identifying the pilot is also very low. As a consequence of these factors, the expected damages cost borne by the drone operator will be a small fraction of the harm caused, and the drone operator will accordingly exercise insufficient care to avoid privacy violations. This can only be an efficient outcome if the cost of reducing or eliminating the uncertainty is very high and there are no other options for protecting privacy, such as destruction of the offending drone (for a discussion of the potential use of ‘violent self-defence’ against drones see Froomkin and Colangelo, 2015).
In sum, there are sufficient uncertainties in the application of the current body of tort and statute that a person upset by unwelcome surveillance cannot be sure of an acceptable resolution, even when that surveillance takes place in a location where they have a reasonable expectation of privacy. From an economic perspective this imposes uncompensated costs on the victim. More importantly, because the drone operator does not face the cost of his or her actions, he or she will not take sufficient precaution to avoid privacy violations and will have an activity level that is too high (Shavell, 1980).
Conclusion
The general public is concerned about the ability of drones to violate personal privacy and surveil activities conducted in spaces where people have a reasonable expectation of privacy. Experimental evidence demonstrates that even individuals who consent to surveillance experience a range of negative emotions, including fear, anxiety and anger, and change their behaviours in response to surveillance. These negative emotions and the behaviour changes are economic costs that must be taken into account when determining the efficient use of drones.
An efficient outcome could in theory be achieved via tort. However, New Zealand’s privacy torts may set too high a standard (‘highly offensive’), and in any event this standard would need to be tested in court to definitively determine what level of drone surveillance meets the threshold. The costs of such action are high, effectively preventing tort from acting as an efficient mechanism for addressing privacy violations. The Privacy Act’s offence of an ‘interference with privacy’ potentially provides a mechanism that more readily facilitates the transfer of cost to the drone operator. However, the privacy commissioner has held that if a drone is not recording imagery then no ‘collection’ occurs, and hence no interference with privacy occurs. This provides the obvious incentive for any drone operator subject to a Privacy Act complaint, but who has not actually published imagery, to simply claim that no information was collected. Additional problems may arise in identification of the drone operator.
This article has identified areas where New Zealand’s current privacy framework requires clarification to better accommodate the challenges posed by drones. Some of the modifications could potentially be achieved by way of a code of practice issued under the Privacy Act, which may provide a relatively low-cost means of setting the standard of acceptable behaviour. Challenges will still remain because the characteristics of drone technology make it difficult to identify the operator, which in turn makes it difficult to obtain any legal remedy. Such challenges may mean that in some instances an alternative, more direct means of intervening to protect one’s right to privacy would be efficient.
sUAS Guide / Q3 Update, October 2016 27