What surveillance means
Surveillance is the systematic observation of people, their movements, communications or behavior. It can be done by governments, businesses, employers, schools or individuals. The word covers a very wide range: a camera watching a shop entrance, a system that reads license plates, software that logs activity on a work computer, and the collection of online behavior by advertisers.
Some forms of monitoring are widely accepted as useful, and some are widely questioned. Many are in between, which is why the topic is debated rather than settled. This article describes common tools and the main arguments, without taking a side on any specific policy.
Common monitoring tools
Video cameras. These are among the most familiar tools, used in shops, transit, streets, homes and workplaces. Recordings can deter or document incidents. Analysis software may add features such as counting people or flagging unusual movement, and some systems apply facial recognition to identify individuals, which is among the most contested uses.
License plate readers. Cameras that read plates can help locate vehicles of interest, but they also create records of where ordinary vehicles have been.
Location tracking. Phones and vehicles generate location data through satellite navigation, mobile networks and Wi-Fi. Such data can support emergency response, but also reveal routines.
Communications monitoring. Governments and companies may be able to access certain communications data under particular legal conditions. The strength of message encryption affects what can be read in transit, and the question of how much access authorities should have is a long-running debate.
Online behavior tracking. Websites and apps record activity, often for advertising, as with online tracking. Some people do not see this as surveillance because it is commercial; others see it as the largest form of monitoring there is.
Workplace and school monitoring. Employers may track computer use, location or productivity, and schools may use filtering and monitoring tools. Related practices appear in algorithmic management, where software supervises work.
Biometric systems. Fingerprint, face and voice checks can identify people. When used to unlock a device, many view them as convenient. When used to identify people in public without their involvement, they raise different concerns. The technology behind biometric authentication differs from its use for identification.
Connected devices. Doorbell cameras, smart speakers and other smart home devices can record in homes and neighborhoods, sometimes capturing people who never agreed to be recorded.
Remote sensing. Drones and satellites can observe from above. Earth observation from orbit is one example.
Data combination. Perhaps the most significant development is not any single device but the joining of records. Separate pieces of information, such as camera footage, purchases and location, can build a detailed picture when combined, and may be bought and sold through data brokers.
Arguments for monitoring
People who support particular tools commonly point to:
- Safety and crime prevention. Cameras and other tools may help investigations and, in some settings, may discourage harm. Evidence on how well they deter crime varies by context and is not uniform.
- Emergency and public services. Location and sensor data can help find people in trouble and manage traffic or infrastructure, as in smart city sensors.
- Fraud and security. Monitoring for unusual behavior can detect account takeovers.
- Convenience. Many services only work because they learn about you.
- Accountability. Recordings can also hold institutions to account, such as body-worn cameras that capture interactions.
Arguments for caution
Critics commonly raise:
- Privacy as a basis for freedom. People who feel watched may change what they say, read or attend, sometimes called a chilling effect. This can matter for lawful activities like political participation.
- Errors and bias. Automated systems can misidentify people, and errors are not always evenly distributed. Algorithmic bias is one way this can arise.
- Function creep. Tools introduced for one purpose can expand to others.
- Power imbalance. The watcher often knows far more about the watched than the reverse.
- Security of the data. Collected information can be leaked or misused.
- Normalization. Gradual introduction may make monitoring seem ordinary before the public has debated it.
Thinking about the balance
The debate is rarely about surveillance versus no surveillance. More often it is about particular uses, and a few questions tend to recur.
- Necessity. Is there a real problem, and is monitoring a reasonable way to address it?
- Proportionality. Is the intrusion matched to the benefit, or would a lighter approach do?
- Transparency. Do people know what is monitored, by whom and why?
- Oversight. Is there independent review, and a way to challenge errors or abuse? Questions of human oversight and who is responsible when an automated system gets it wrong apply here too.
- Limits on retention and sharing. How long is data kept, and who can access it?
- Consent and choice. Can people opt out, or is participation effectively required?
Different societies answer these differently, and laws vary widely. People in the same community can reasonably disagree depending on what they value and what they trust.
What individuals can do
Individuals cannot decide public policy alone, but they can learn what monitoring exists where they live and work, read the notices posted, use available privacy controls, and take part in public consultations where they exist. At home, thoughtful placement and configuration of cameras and microphones, and telling visitors about them, respects others’ privacy.
No one can say in general where the right balance sits. What is clear is that the technology is becoming cheaper and more capable, which makes it more important that the questions above are asked early rather than after the tools are everywhere.