Narrative Goldmine

Personal Data Store

9 min read

Properties

Type
  • Class
Resource
  • urn:ngm:class:personal-data-store
Domain
  • data
Status
  • stable
Maturity
  • emerging
Is a
Requires
Enables
Generated
  • by: process:vault-migrate/1.0 · at: 2026-09-22T12:41:39.024503178Z
Bridges to
Contrasts with
Legacy term id
  • MV-9184
Links
    Page resource
    • urn:visionflow:page:161b6fdb3fe1beccf8016bd942f46b9a589b69a39aa1c74de31881441fe54d44
    Related to
    Standardized by
    Supports
    Uses

    A Personal Data Store (PDS) is a user-controlled software system or vault that aggregates, stores, and selectively discloses an individual’s personal data — including identity attributes, health records, behavioural logs, and transaction histories — under the direct control of the data subject. It enforces consent-based access through verifiable credentials and cryptographic authorisation mechanisms, enabling fine-grained, revocable sharing with third-party services. PDSs embody the principles of data sovereignty and privacy-by-design, decoupling data custody from service providers and restoring ownership to individuals. They serve as a foundational layer for interoperable digital identity ecosystems, federated personal AI assistants, and regulatory compliance with frameworks such as GDPR and CCPA.

    Overview

    • The Personal Data Store concept emerged from the recognition that the dominant model of platform-controlled data silos — in which user data is harvested, processed, and monetised by service providers without meaningful user control — is both privacy-damaging and economically inefficient.
    • A PDS inverts this model: instead of data flowing to and residing inside each application’s private database, it remains in a single user-controlled location and applications are granted time-limited, revocable read or write access.
    • This paradigm shift is sometimes described as moving from “data as a byproduct of service use” to “data as a personal asset under individual stewardship.”
    • The concept intersects with Self-Sovereign Identity (SSI), where individuals hold cryptographic keys that prove ownership of identity attributes without relying on central directories.
    • Architecturally, a PDS typically consists of:
      • A secure storage layer (local device, personal cloud, or federated server)
      • An Authentication and Authorisation framework (OAuth-2-style grants, DID-based auth)
      • A data schema layer (often Linked Data vocabularies such as schema.org, FHIR for health, vCard for contacts)
      • A consent dashboard that surfaces what data exists, who can access it, and how to revoke access
      • APIs or protocol adapters (Solid, ActivityPub, OIDC, DIDComm) for application interoperability
    • Maturity is currently rated “emerging”: several implementations are production-grade but mass-market adoption remains limited pending ecosystem coordination.

    Key Components

    • Storage backend: encrypted object store or graph database hosting the individual’s data under user-controlled keys; may be self-hosted, hosted by a trusted provider, or distributed across edge devices.
    • Decentralized Identifiers (DIDs): provide globally unique, cryptographically verifiable identifiers for the data subject that are not dependent on any central registry.
    • Verifiable Credentials: W3C-standardised data structures enabling third parties to issue attestations (age, qualifications, health status) that the PDS holder can selectively present without revealing the underlying raw data.
    • Access Control policies: fine-grained rules specifying which agents, for which purposes, for which time windows, may read or write which data subsets; implemented as Web Access Control (WAC) in the Solid ecosystem or as ZCAP-LD capability tokens.
    • Consent Management layer: user-facing interface and machine-readable policy store tracking consent receipts as defined by ISO/IEC 29184 and the Kantara Consent Receipt specification.
    • End-to-End Encryption: data at rest and in transit is encrypted using keys held only by the data subject, preventing even the PDS hosting provider from accessing content in plaintext.
    • Data Portability APIs: standards-compliant export and import interfaces (e.g., GDPR Article 20 mandated machine-readable export, DTP formats) ensuring the user can migrate without lock-in.
    • Linked Data schemas: RDF-based vocabularies that make stored data semantically interoperable across applications without requiring bespoke integration work per data pair.
    • Audit log: immutable, append-only record of all data access events, allowing the owner to retrospectively verify whether access terms were honoured.
    • Agent interface: programmatic APIs enabling trusted Personal AI Assistant agents to query the PDS on behalf of the user, e.g., to provide personalised recommendations without exporting data to cloud AI services.

    Applications and Use Cases

    • Healthcare data control: a patient’s PDS aggregates EHR records from multiple providers; the patient grants a specialist temporary read access to relevant records only, revoking it after the consultation. Compatible with FHIR data standards.
    • Financial data aggregation: under Open Banking (PSD2 in the EU, CDR in Australia), a PDS acts as a consent broker, pulling transaction history from banks and selectively sharing it with budgeting apps or mortgage lenders without intermediary data brokers.
    • Personalised advertising on user terms: instead of tracking cookies and third-party data brokers, advertisers query a user’s PDS for declared preferences; the user is compensated or simply retains control over what is shared via a Data Marketplace model.
    • Education credential wallet: academic certificates issued as Verifiable Credentials are held in the PDS; the holder presents cryptographic proofs of qualification to employers without revealing other transcript details.
    • Metaverse identity portability: avatars, reputations, and in-world asset ownership are anchored to a PDS-held DID, allowing the user to carry their digital identity and social graph across virtual environments without platform re-registration. Connects to Federated Identity patterns in the metaverse.
    • Federated machine learning: a research consortium’s Federated Machine Learning pipeline queries consented data subsets from participant PDSs; raw data never leaves the individual’s store, enabling population-level insights with individual privacy preservation.
    • Smart-city services: municipal apps (transit, waste collection, local alerts) read only the attributes a citizen has explicitly shared from their PDS, avoiding the surveillance model of centralised city data lakes.
    • Decentralised social networking: ActivityPub-connected PDSs let users host their own social graph data, enabling cross-platform following and posting without dependence on a centralised platform’s terms of service.

    Standards and Context

    • W3C Solid: Tim Berners-Lee’s open standard (originally developed at MIT, now at Inrupt and standardised at W3C) that specifies a protocol for personal online datastores (Pods) using Linked Data principles. Solid Pods are the most mature open-standard PDS implementation available today.
    • W3C DID Core: defines Decentralized Identifiers, the identity anchor mechanism for most PDS architectures, allowing authentication without centralised registries.
    • W3C Verifiable Credentials Data Model: specifies the credential format for attestations issued to and stored within a PDS.
    • GDPR (General Data Protection Regulation, EU 2016/679): Articles 15–20 (rights of access, rectification, erasure, restriction, portability, and objection) create a regulatory mandate for PDS-like capabilities, driving adoption in European markets.
    • CCPA / CPRA (California Consumer Privacy Act / Privacy Rights Act): analogous US state-level regulation driving similar data rights in the American market.
    • ISO/IEC 29184: international standard for online privacy notices and consent, relevant to PDS consent-receipt design.
    • Kantara Consent Receipt: community specification for a machine-readable JSON consent receipt, commonly referenced in PDS consent layers.
    • MyData Global: non-profit coalition promoting human-centric data management; the MyData Operator certification scheme provides a governance framework for PDS hosting services.
    • OpenID Connect (OIDC) and OAuth 2.0: widely used protocol layer for delegated authorisation flows between PDS and third-party applications.
    • DIDComm: DIF (Decentralised Identity Foundation) messaging protocol enabling secure, private communication between DIDs; used in PDS-to-agent and PDS-to-PDS protocols.
    • ActivityPub (W3C): federated social networking protocol increasingly paired with PDS architectures to enable decentralised social graphs.
    • Regulatory tailwinds (GDPR, DSA, Data Act in the EU; CDR in Australia; emerging US federal privacy bills) are accelerating PDS adoption by mandating data portability and right-to-erasure capabilities that central silos struggle to implement cleanly.

    Contrasting Concepts

    • Centralised Data Silo: the dominant alternative, where each service platform holds user data in its own proprietary database with limited user visibility or control.
    • Data Broker: commercial entities that aggregate personal data from many sources and sell it; PDSs disintermediate this model by returning data ownership to the individual.
    • Cloud storage services (Google Drive, iCloud): offer user-accessible storage but do not implement consent-based third-party access controls or semantic interoperability at the PDS level.

    Current Landscape (2026)

    • Solid, the leading personal-data-store (Pod) architecture, moved onto the formal W3C standards track in October 2024 with the new Linked Web Storage Working Group (co-chaired by Inrupt’s Aaron Coburn), graduating the protocol from Community Group draft to an official work item alongside HTML and Verifiable Credentials.
    • The Open Data Institute (ODI), co-founded by Tim Berners-Lee and Nigel Shadbolt, took over stewardship of the Solid project, protocol and community from October 2024, folding it into its data-stewardship portfolio.
    • The frontier has shifted decisively towards AI: Inrupt released Enterprise Solid Server (ESS) 3.0 in mid-2026 (“making consumer data ready for AI assistants”), adding a standards-based MCP server bridging AI agents to consented Pod data, native identity-provider support (Okta, Azure AD, Ping), stable identifiers and environment isolation.
    • “Agentic Wallets” emerged as a design pattern combining Solid’s Web Access Control with Anthropic’s Model Context Protocol to enforce least-privilege, time-bound agent access; the academic Liquid Protocol (OpenReview, 2025) formalised MCP-to-Pod operations, arguing PDS query-at-source enables instant GDPR Right-to-be-Forgotten compliance versus data baked into model weights.
    • Regulation converged on user-held stores: eIDAS 2.0 (Regulation (EU) 2024/1183) mandates every Member State offer a European Digital Identity Wallet by end of 2026, with a further tranche of implementing regulations (EU 2026/1730, 2026/1731, 2026/1735) published 22 July 2026 specifying PID formats (ISO mdoc, SD-JWT VC), selective disclosure and relying-party registration.
    • The EU Data Governance Act’s data-intermediation rules became applicable in September 2025 and most Data Act provisions from 12 September 2025, but uptake of registered Data Intermediation Service Providers stayed low; MyData Global’s October 2025 “Human-Centric Roadmap for Europe” flagged that DMA real-time data portability remains largely theoretical and proposed a lighter-touch DIASP category plus a “European Personal Cloud” programme.
    • Consolidation continued around consumer reach: Project Liberty’s Frank McCourt confirmed Solid-integration talks with Berners-Lee at SXSW 2025, while Inrupt’s own sandbox Pods still discourage storing real personal data, underlining the open challenge that a mature, paid, consumer-grade Pod-hosting market has yet to materialise.

    References

    Provenance

    Graph View

    Backlinks