3 identity levels
Identity Levels
An identity level is how far an instrument resolves the person behind what it observes: whether it cannot tell one person from another, can tell them apart without knowing who they are, or knows who they are. It is the instrument's, never the event's; the same visit is anonymous to one instrument and pseudonymous to another.
What you can do with a identity level
Its identity levels— 3
- AnonymousThe instrument cannot tell one person from another: an impression, a public conversation, a cost, an aggregate reading.
- PseudonymousThe instrument can tell one person from another without knowing who they are: a device, a cookie, a click identifier, a session.
- KnownThe instrument knows who the person is: an account, an email address, a phone number, an order, a consent record.
Why it matters
What goes wrong without it, as the commitments that license this kind say it, each standing on its sources.
- A person is known at a declared levelA journey machine moves a person, and a move counted by an instrument that cannot tell one person from another is a move nobody made: an impression cannot move anyone, however many there are. The profile standards keep one representation of a person across the identifiers each source knows them by, and the regulation keeps pseudonymised data apart from identified data by exactly the additional information it takes to join them; a model that declares grain on its events but not resolution on its instruments cannot say which of its readings are about people at all. Declaring the level on the instrument, where the identifier lives, is what lets the build hold every move to evidence that could have been one person's.
XDM Individual Profile Class · Regulation (EU) 2016/679, Article 4: Definitions · ISO/IEC 24760-1:2019 IT security and privacy: a framework for identity management, Part 1: Terminology and concepts