home / kskb / notes

notes: research/papers/kerione-quantum-secure-dids

This data as json

path title type status classification words body
research/papers/kerione-quantum-secure-dids Quantum Secure DIDs summary v0.1 — 2026-10-14: drafted by Claude from the PDF; unread by Mark; marked partner on Mark's word 2026-10-21, paraphrase only partner 589 # Quantum Secure DIDs **Abstract.** A Rebooting the Web of Trust (RWOT) community paper by Carsten Stöcker, Samuel M. Smith and Juan Caballero (version 1.0, July 2020) on how to make a decentralized identifier (DID) method resistant to quantum attack using existing signing and hashing primitives. It proposes publishing only hashes of public keys and rotating keys after every signing operation, and points to the pre-rotation in Key Event Receipt Infrastructure (KERI) as one way to do so. Written for designers of DID methods and decentralized identity systems. ## Summary The paper starts from the risk that quantum computers may break the signature schemes on which decentralized identity relies (elliptic-curve and RSA-based schemes), with no warning of when. Its objective is a mechanism, independent of any one DID method, for protecting DIDs during the transition to quantum-resistant signing ciphers, using today's ciphers. Its prior-art section rests on Daniel J. Bernstein's cost analysis: quantum computers are expected to be far more effective at factorization than at finding hash collisions. The authors conclude that existing signing primitives should be treated as vulnerable while hash functions can be treated as resilient. The paper takes its inspiration from Bitcoin's pay-to-public-key-hash addresses, where the public key is hidden behind a hash until first use. It notes that did:ethr and did:sov use similar hash-based blinding, but that the public key is exposed once a DID document is configured or a verifiable credential is signed. The proposed measures are to register one-way hashes of public keys rather than the keys themselves as authorization attributes, and to deactivate or rotate a key every time it is exposed by signing. Because rotating on a blockchain after every signature would be costly, the authors suggest moving key-state maintenance to a lighter parallel system such as KERI, whose verifiable logs of key events do not depend on block confirmation times. A "recipe" follows as three user journeys: initializing a DID from high-quality randomness with keypair tuples, signing with one key of a tuple while using the other to sign a rotation, and verifying a credential against the key's validity and deactivation times. The KERI section explains pre-rotation: each rotation commits to a digest of the next key set, which the authors describe as post-quantum secure. The conclusion proposes a more rotation-intensive paradigm for high-stakes DIDs and suggests the method could become a non-normative addendum to a KERI reference implementation. ## Key terms DID, key rotation, one-way hash function, pre-rotation, keypair tuple, pay to public key hash (p2pkh), Key Event Receipt Log (KERL), post-quantum security, verifiable credential ## Claims and their evidence - Quantum computers will be far more efficient at large factorizations than at collision searches, so hash functions remain resilient while current signature schemes do not. [informative] (resting on Bernstein's cited analysis) - An unused Bitcoin p2pkh address can be considered quantum secure because the public key has not been revealed. [informative] - Registering public-key hashes and rotating after each exposure makes a DID method much more quantum-secure with existing primitives. [informative] - KERI's pre-rotation, with the next key set committed as a digest, can be considered post-quantum secure. [informative] - The mechanism is feasible today on current DID systems and a KERI reference implementation. [unverified] ## Related The KERI white paper (Key Event Receipt Infrastructure, Smith, arXiv 2019); Decentralized Autonomic Data (DAD) and the three R's of Key Management; A DID for Everything. ## Provenance Drafted by Claude on 2026-10-14 from the PDF (10 pages), read in full apart from the reference list. Not yet read by Mark.

Links from other tables

  • 1 row from summary in resources
Powered by Datasette · Queries took 3.264ms