---
name: Privacy zuckering
slug: privacy-zuckering
category: pattern
status: published
created: 2026-08-21T00:00:00.000Z
modified: 2026-08-21T00:00:00.000Z
definition: Steering people into sharing more about themselves than they meant
  to, by making the permissive choice the easy one and the private choice
  several clicks deep.
aliases:
  - name: privacy dark pattern
    source: community
  - name: oversharing by default
    source: community
  - name: consent maze
    source: community
tags:
  - consent
relations:
  contrastWith:
    - obstruction
  variantOf:
    - dark-pattern
  partOf: []
  seeAlso: []
implementations: []
sources:
  - title: "Deceptive Design: types of deceptive pattern"
    url: https://www.deceptive.design/types
demo: inline
exhibit: false
useWhen: the settings are arranged so sharing is the easy answer
---

Privacy zuckering is named after Mark Zuckerberg, coined by the Electronic Frontier
Foundation in 2010 for the way Facebook's settings kept resolving in favour of sharing,
and it entered Harry Brignull's original catalogue of twelve deceptive patterns. Its
current status is worth stating plainly, because the taxonomy moved: the types index at
[deceptive.design](https://www.deceptive.design/types) no longer lists it. The behaviour
it described has been sorted into the mechanism-named types that replaced it, mainly
sneaking, obstruction and preselection, which is a better basis for enforcement than a
person's surname. The word survives in critique and in writing about privacy because it
names a *domain* rather than a mechanism: whatever the individual trick, the thing being
extracted is personal data.

The mechanism is an asymmetry of effort, not a lie. Both choices are present, both are
reachable, and nothing on screen is false. One of them is preselected, sits at the top,
carries the word "recommended", and needs no clicks (see
[preselected opt-in](/preselected-opt-in)). The other is behind "More options", one level
down, in a list that has to be read. Then the confirm button is labelled for the flow
rather than for the outcome: "Continue" and "Get started" commit the reader to a sharing
setting without naming what is being shared, which is the
[misdirection](/misdirection) half of the pattern working on the same screen as the
[obstruction](/obstruction) half. Effort is the tax, and it is levied only on privacy.

The test for a design you are building is symmetry, and it is easy to apply. Can the
private choice be made in the same number of interactions as the permissive one, at the
same reading level, with the same visual weight? Does the button that commits the choice
name the outcome ("Share with everyone") rather than the transition ("Continue")? Is the
setting still what the reader chose a month later, after an update that "improved" the
privacy controls? A flow that fails the third question is a
[roach motel](/roach-motel) as well: easy in, hard out. The related family case is
[friend spam](/friend-spam), where the data extracted is not even the reader's own but
their address book, and the consent covers looking rather than sending.

There is a legal edge to this now that did not exist when the term was coined. Consent
under GDPR has to be freely given, specific and informed, and regulators have been
explicit that a design where accepting takes one tap and refusing takes five is not
freely given. That is why cookie banners in Europe grew a "Reject all" button of equal
prominence beside "Accept all", and why the same reasoning applies to profile visibility,
location history and ad personalisation. As a general [dark pattern](/dark-pattern) it is
also simply bad business: the settings people did not choose are the settings they
disable in a rush the first time a story about the product breaks.