Most of today’s video surveillance cameras are capable of providing audio, but when should it be implemented? Acceptable uses of audio surveillance, as opposed to video, are a different concept in many jurisdictions worldwide. Privacy laws regarding audio may be more stringent than those for video, but both tend to centre around the concept of an “expectation of privacy,” which may mean something different in the case of audio versus video. How can system designers and end users leverage the benefits of audio surveillance without violating privacy? Specifically, we asked this week’s Expert Panel: Do privacy concerns limit the use of available audio features on video cameras? What are acceptable uses of audio in surveillance systems?
Video surveillance has become the norm, accepted by the general population of most countries. Audio surveillance is another story, and privacy is at the root of the issue. Regardless of privacy laws, the general population is uncomfortable with their private conversations being recorded without their knowledge. A few exceptions exist. When the area being recorded is posted with appropriate signage, there is more acceptance.
There are a lot of security applications where the addition of audio to the video enhances security. But an important thing to note is that audio privacy laws are very different from those for video. According to the Electronic Communications Privacy Act of 1986, legally in the United States, there is an “expectation of privacy” – i.e., when people expect to have a private conversation, monitoring is prohibited. Conversely, when there is no expectation of privacy – e.g. in public spaces where conversation could easily be overheard – audio monitoring may be allowed. Businesses can remove the expectation of privacy with a sign saying that audio monitoring is taking place. Apart from the federal laws, there are also differences among states. In some instances, using audio may be acceptable, such as if there is a known threat or incident or when the audio level goes beyond a threshold (e.g., breaking a window).
Audio functionality has long been a feature of physical security systems starting with two-way audio or “listen in” capabilities offered by a major systems integration franchise many years ago. And it’s also been a common feature in video surveillance cameras, with many units having built-in microphones. Audio brings another facet to security, allowing operators to hear what’s going on at the protected premises under alert or alarm. Some, with two-way audio, permit a live, two-way conversation. However, as with any kind of surveillance technology, there may be limitations with regard to privacy concerns. As many know, recording conversations requires consent. However, in areas where there is no reasonable expectation of privacy, audio recording or live audio may not be a problem. It’s definitely a balancing act, and systems integrators and value-added resellers should always check local laws and regulations and consult with legal counsel if in doubt.
Privacy restrictions are important and must exist, and perhaps ought to be seen as a condition of the use of video surveillance rather than a limitation placed upon the industry. In some cases, audio can be more intrusive than video. With the audio capabilities of most of today’s IP cameras, it might be tempting to complement video recording with sound. However, in most public places, approval to do so is required from local or other authorities. Any use of video security systems, with or without audio, must be carefully considered prior to deployment to make sure privacy remains uncompromised. Permits to deploy cameras will not, by default, include audio permissions unless those who are being recorded explicitly give consent. In most countries, recording equipment must be advertised if it is in use, i.e. clear signs must be posted informing people in the area that recording equipment is in use.
In the United Kingdom, CCTV is subject to the Data Protection Act and Code of Practice by the Information Commissioner’s Office, revised in May 2015. It now addresses audio, particularly because of proliferating body-worn video. Ideally, it should be independent of recording video. For instance, audio may be too intrusive if it hears outside the camera’s field-of-view. Conversely, audio may be better when video embarrasses a partially clothed subject. Audio without continuous recording may raise alarms, e.g. gunshot or shouting. Public help-point intercoms should only be initiated by a caller and not permit controllers to listen-in generally. In fact, general recording in taxis is outlawed. Like video, the audio must be “accurate,” which means sufficiently clear. How reliably can someone be identified by their sound? Importantly, the public under surveillance must be informed it is happening and who holds it.