v0.1.0
Protocol Spec

DTI Alignment

How PDPP relates to the Data Transfer Initiative and fits into the broader data portability landscape.

DTI's receptiveness to external proposals

  • DTI actively solicits external input. CLA-based contribution to DTP codebase. Membership not required.
  • Chris Riley led a Brussels working session (early 2024) to gather external input on the third-party trust model.
  • 2025 Annual Report: goal is to "be at the table whenever data portability is on the agenda."
  • May 2025 blog: "trust is built not on technology alone, but on people and institutions."

The gap we fill

DTI has publicly acknowledged:

  • "Unclear consent processes" in current portability frameworks
  • Standard OAuth scopes are too broad for continuous/real-time portability
  • EU DMA mandates "continuous and real-time" portability but doesn't specify how third parties should be authorized
  • Missing infrastructure for fine-grained, parameterized access controls

No one has proposed parameterized OAuth grants (RFC 9396-based) to DTI. This has only been used in Open Banking.

What resonates with DTI

  1. Solving the DMA "continuous and real-time" mandate with scoped, ongoing grants
  2. Trust and harm mitigation: parameterized grants prevent over-sharing
  3. User empowerment through explicit, understandable consent
  4. Grounded in standard web protocols (OAuth, RFC 9396)

What to avoid

Chris Riley explicitly criticized "short-lived dreams of 'trustless' technology, powered by blockchains and math." Keep the proposal grounded in institutional trust, not cryptographic trustlessness. Vana's blockchain/Web3 aspects should not be the lead framing for DTI engagement.

Path to engagement

DTI is an independent 501(c)(4) nonprofit (not Linux Foundation). Path is direct engagement:

  1. Build the spec and working implementation
  2. Publish openly
  3. Engage Chris Riley / DTI team directly
  4. Propose as a contribution to the DTP ecosystem for fine-grained consent

Spec implications

The OAuth approach (define the grant, not the resource API) is the right framing for DTI. The spec should:

  • Use RFC 9396 as the envelope (standard web protocol, not custom)
  • Define the grant as a portable consent artifact that any server can enforce
  • Define the connector protocol as a reference implementation for data collection
  • Explicitly not define the resource API (like OAuth doesn't define what you do with the token)
  • Frame continuous/recurring grants as the answer to DMA's real-time portability requirement

Source: Gemini 3.1 Pro Preview research with Google Search (2026-03-28), checking dtinit.org, conference talks, blog posts, governance docs.