A wallet pseudonym is a stable identifier a holder presents to one relying party. By default the wallet never hands the same pseudonym to a second relying party, and colluding relying parties cannot tell two pseudonyms belong to one person. It lets a service recognise a returning user without learning who they are.
ARF termReviewed 7 October 2026
Between the wallet and the relying party, when the user shares data.
Issuer
Signs credentials: a member state, university, bank or employer
Issuance
EUDI Wallet
Belongs to the user, who decides what to share
Presentation
Relying party
The service that requests and checks the data
Trust layer: who is allowed to sign, and is it still valid?
Rules: the law and technical framework behind it all
The wallet derives a different pseudonym per relying party. Your service sees the same value each time that person returns, so accounts, preferences and reputation persist. Two services comparing notes still cannot tell they are looking at the same human.
This answers the usual objection to identity systems, which is that one strong identifier everywhere makes tracking across services trivial. Per-verifier derivation removes the shared key that would allow it.
The regulation backs this up. Article 5b(9) says a relying party may not refuse a pseudonym where law does not require identification. The ARF also describes scope rate-limited pseudonyms, which cap how many a user can hold for a given scope, and setting that rate to one gives you one account per person. Treat that last part as a direction rather than a feature: ARF v3.0.0 specifies the requirements such a protocol must meet but does not yet specify the protocol itself.
Defined in
ARF v3.0.0 sections 2.5 and 4.7, Annex 2 Topic 11; CIR (EU) 2024/2979 Art. 14(2); eIDAS 910/2014 Art. 5b(9)
Uniqueness and identity are different requirements, and most platforms only need the first. Pseudonyms let you enforce one account per person without becoming the custodian of an identity database.
