Privacy by Design
Most privacy advice puts the work on individuals: change your settings, read the terms, opt out of sharing you never opted into.
Privacy by design is an approach to building technology in which protection is part of the system from the start, not added later or left to the user. The private choice is the default choice.
The common question is, "Did this person configure the tool correctly?" The more useful one is, "Why did the tool need configuring to be safe?"
Cars work this way. Nobody is asked to install their own seatbelts or wire in the airbags. The protection comes with the car, and you're safer whether or not you ever think about it.
Defaults decide more than settings
Most people never change a default. That is a reasonable choice: people use a tool to get something done, and auditing it is a separate job. So the default is the policy for nearly everyone, and when safety depends on settings, it ends up belonging to the people with the most time, confidence, and technical knowledge. Why Most People Never Change the Default explains why defaults stick and who ends up deciding.
Ask: what happens to people who never open the settings?
A promise is not a guarantee
There is a difference between a company that promises not to look at your data and a system that cannot. A promise can change when a policy is rewritten, a company is sold, or a court orders access. A guarantee, such as end-to-end encryption, is built into the system, though it has limits of its own. Is It a Promise or a Guarantee? walks through both with real cases.
Ask: when a tool says it respects privacy, is that a promise or a guarantee?
Collect less
The most reliable protection for data is not holding it. Data minimization means collecting only what a service needs to work and deleting it when it is no longer needed.
Data that was never collected cannot be breached, sold, subpoenaed, or repurposed. Every extra field a vendor keeps is a future risk carried by the people it describes, usually without their knowledge.
A small example: a sign-up form asks for a child's full birthdate when all it needs to know is whether they're over thirteen. The extra detail does nothing for the service, and it sits in a database for years.
Ask: what is this tool collecting that it doesn't need to do its job?
It is also social design
Privacy by design is usually discussed as engineering: encryption, retention limits, protective defaults. The same logic applies to how communities and institutions adopt tools.
Building codes are a good comparison. Tenants aren't expected to inspect the wiring before they move in. The code makes the safe choice the standard choice, decided once, upstream, by people responsible for it.
A district that negotiates data limits into a vendor contract is designing privacy. So is a community that chooses tools where members do not need to become security experts to take part. In both cases, the goal is to reduce the number of decisions a person must get right to stay protected.
Ask: who in our group is deciding the defaults for everyone else?
Start close to home
Try the questions on one real case: the next app your school, workplace, or community group asks you to use, or asks you to make others use.
- What does it do with data if no one changes anything?
- When it says it protects privacy, is that a promise or something it technically cannot do?
- What does the sign-up ask for that the app doesn't need?
- If it is breached, who carries the risk: the company, or the people in its database?
Limits and open questions
Protection and usability can pull against each other. The most secure tool is not always the easiest to use, and some of the easiest tools achieve ease by collecting more data. Privacy by design does not remove these tradeoffs. It asks that they be made deliberately, by the people building and choosing the system, instead of being passed silently to everyone who uses it.
Defaults can also be paternalistic. A protective default still needs to be understandable, and people need a way to make a different choice when their circumstances call for it. A good default should make a choice easier to understand, not invisible.
Open questions:
- Who should set defaults for a community: the people who run the tools, or the people who use them?
- When a protective default makes a tool harder to use, how much friction is worth it?
Newsletter connections
- Nobody Changes the Default: the largest child-safety settlement in tech history barely moved the defaults that matter.
- Looks Good on Paper: the protection written down is not always the one people live with.
- Who Said Yes?: millions of students' data exposed by a testing vendor no family chose.
- Attack or Outage?: personal security habits matter, but the software behind freezers and water pumps has to be secured by the organizations that run it.
Related ideas
- What Can People Do About AI?: four questions for finding who decides, which is often the same as who set the default.
- Why Most People Never Change the Default: why whoever sets the default makes the decision for nearly everyone.
- Is It a Promise or a Guarantee?: the difference between a company that promises not to look and a system that cannot.
Appears in
- Privacy: the first stop, the core idea the route builds on.
Sources and further reading
- Ann Cavoukian, Privacy by Design: The 7 Foundational Principles (Information and Privacy Commissioner of Ontario, 2009).
- General Data Protection Regulation, Article 25: Data protection by design and by default.
What changed
- 2026-11-19: First published.
Spot something that needs correcting or updating, or have an example from your own community? Email me at hello@wiobyrne.com.