Privacy is hard to define but easy to miss. Most people would object if a stranger read their diary, followed them around town or listed their medical conditions online, even if nothing illegal happened. Digital technology has made collecting, storing and combining personal information cheaper than ever, which makes the old questions urgent.

What privacy protects

In 1890 the American lawyers Samuel Warren and Louis Brandeis, worried about newspaper gossip and the new portable camera, argued for a legal right of privacy, which they described as “the right to be let alone.” Later thinkers stressed control over information about oneself. The philosopher Helen Nissenbaum proposed the idea of contextual integrity: information flows carry norms from the context where they occur, and privacy is violated when information moves in ways that break those norms. Your doctor knowing your diagnosis is fine; your employer learning it from a data sale is not, even though the information is the same.

Privacy matters for more than individuals. People who feel watched tend to censor themselves, a chilling effect that can weaken free speech, journalism and political organizing.

Data brokers

Data brokers are businesses that collect personal information from many sources, such as public records, purchases, apps and websites, and sell or license it to others, usually without any direct relationship with the people described. Their profiles can include addresses, income estimates, interests and health-related guesses. They are used for marketing, fraud prevention, background checks and more. Critics worry about accuracy, about sensitive data such as location being sold, and about how hard it is for people to learn what is held about them.

Surveillance capitalism

The term surveillance capitalism was popularized by the scholar Shoshana Zuboff in her 2019 book The Age of Surveillance Capitalism. In paraphrase, she used it for business models that treat data about people's behavior as raw material for predicting and influencing what they will do, and that sell those predictions, often to advertisers. Supporters of the term say it captures something new. Critics reply that targeted advertising pays for many free services, that its persuasive power may be overstated, and that the term lumps very different practices together.

Government surveillance

Governments collect information for security, law enforcement and public services. In 2013 documents leaked by the contractor Edward Snowden revealed that the US National Security Agency had been collecting records of Americans' phone calls in bulk. Defenders argued that such programs help prevent terrorism under legal oversight; critics argued that they swept up innocent people and lacked real checks. In 2015 the USA FREEDOM Act ended that bulk collection of phone records. The underlying question remains: how much security is worth how much loss of liberty, and who keeps watch over the watchers?

Consent and its limits

Much data collection rests on notice and consent: a company posts a privacy policy and users click “I agree.” A 2008 study estimated that reading every privacy policy a typical American encountered would take about 76 working days a year. Consent given without understanding or real choice is weak consent.

GDPR

The European Union's General Data Protection Regulation (GDPR), which took effect in May 2018, is the most influential privacy law so far. It applies to organizations anywhere in the world that handle personal data of people in the EU. It requires a lawful basis for processing, gives people rights to access, correct and in some cases erase their data, and allows fines of up to 4 percent of a company's global annual turnover. Many other countries and several US states have since passed laws inspired by it.

Worked example: the running app's map

Imagine a popular running app that records where users jog and, by default, publishes an anonymous heat map of popular routes worldwide. Something like this really happened: in 2018 such a map revealed the outlines of remote military bases, because soldiers jogged inside them. Let us analyze a version of the case.

Contextual integrity. A runner shares her route with the app to track her training, and perhaps with friends. Those are the norms of the context. Publishing patterns that reveal where she lives, or where a secret base is, moves the information into a new context with different norms. No single data point seems sensitive; the combination is.

Consent. The app's policy mentioned the heat map, and a setting allowed users to opt out. But few people read the policy, and the default was to share. Defaults matter because most people never change them.

“I have nothing to hide.” A user might shrug. Privacy scholars reply that privacy is not only about hiding wrongdoing. A survivor of domestic abuse, a journalist protecting a source or an ordinary person who runs alone at dawn all have good reasons to control who knows where they go.

GDPR. For users in the EU, the company needs a lawful basis for processing location data, must explain its use clearly, and must let users access and delete their records. Regulators expect privacy to be built into the design, not added later.

A better design might make public sharing opt-in, hide the start and end of routes near homes, and leave sensitive areas off public maps.