Global Digital Identity Schemes: Technical & Historical Analysis

Date: 2026-01-08 Author: Civ Implementation Team

1. Introduction

This report analyzes the technical specifications and historical evolution of major digital identity schemes globally. The goal is to provide architectural guidance for the civ package, ensuring it supports current standards and future cryptographic transitions.

2. Classification of Identity Schemes

We categorize national ID cards into three primary groups based on their technical architecture.

2.1. ICAO 9303 Compliant (ePassports & Derived IDs)

Scope: Worldwide Passports, UN Laissez-Passer, EU National ID Cards (since 2019), Hong Kong (ePassport app).

• Structure: Logical Data Structure (LDS) with Data Groups (DG1-DG16).
• Access Control:
  • BAC (Basic Access Control): Legacy, uses 3DES/SHA-1 derived from MRZ.
  • PACE (Password Authenticated Connection Establishment): Modern standard, uses ECC/AES with MRZ/CAN/PIN. Mandatory for EU eIDs.
• Interoperability: High. Standardized APDUs (SELECT, READ BINARY with SM).

2.2. ISO 7816 File-System Based (PKI Smart Cards)

Scope: Japan (JPKI/JPDL), USA (PIV), Estonia (EstEID).

• Structure: Hierarchy of Dedicated Files (DF) and Elementary Files (EF).
• Access Control:
  • PIN: Primary method (e.g., JPKI 4-digit PIN, PIV PIN).
  • Certificates: X.509 certificates stored in EFs for offline validation.
• Interoperability: Moderate. Conforms to ISO 7816-4 but AID and File IDs are country-specific.

2.3. Proprietary / Legacy Architectures

Scope: Thailand (National ID), Malaysia (MyKad), China (Resident ID).

• Structure: Often flat memory models or proprietary applets.
• Access Control:
  • Proprietary Commands: e.g., Thai ID uses fixed offsets; MyKad uses SET LENGTH.
  • Closed Systems: China requires a proprietary SAM (Secure Access Module) to decrypt data.
• Interoperability: Low. Requires reverse-engineered or vendor-specific drivers.

---

3. Historical Evolution of Cryptography

The transition of cryptographic standards in ID cards reflects the global race against increasing computing power and cryptanalysis.

Phase 1: RSA 1024 / TDES (2000s - Early 2010s)

• Standards: RSA 1024-bit for signatures, 3DES for encryption. SHA-1 for hashing.
• Japan (Juki-Card): Basic Resident Registry Card (valid until 2025) used RSA 1024.
• Estonia (EstEID): First generation (2002-2011) used RSA 1024.
• Malaysia (MyKad): Introduced in 2001, updated to 64kb PKI in 2002.
• ICAO: BAC (Basic Access Control) established using 3DES.
• Status: Deprecated / Insecure.

Phase 2: RSA 2048 / SHA-256 (2010s - Present)

• Standards: RSA 2048-bit becomes the baseline. SHA-256 replaces SHA-1.
• Japan (JPKI): Current My Number Card (since 2016) uses RSA 2048.
• Estonia (EstEID): Migrated to RSA 2048 in 2011. (Note: 2014-2017 Infineon chips suffered ROCA vulnerability).
• USA (PIV): Mandated RSA 2048 (or higher).
• Status: Current Standard, but facing deprecation by 2030 (NIST).

Phase 3: ECC / AES (Late 2010s - Present)

• Standards: Elliptic Curve Cryptography (P-256, P-384, Brainpool). AES-128/256 for encryption.
• Japan (Next-Gen JPKI): Planned for 2026 introduction, aiming for mass adoption by 2030. Will use ECC P-384.
• Estonia (EstEID): Transitioned to ECC P-384 in 2018 (post-ROCA).
• ICAO: PACE protocol uses ECC (ECDH) for strong session keys. Mandatory for EU eIDs from 2021.
• USA (PIV): Prefers ECC P-256/P-384 for new authentication keys.
• Status: Recommended for performance and security.

Phase 4: Post-Quantum Cryptography (Future)

• Standards: Lattice-based cryptography (Kyber/ML-KEM, Dilithium/ML-DSA).
• Germany: Conducting PQC PoC for eID cards (2024+).
• USA (PIV): NIST and DHS are actively piloting PQC for PIV cards (FIPS 204/205 integration).
• Challenge: Large key sizes require Extended Length APDUs and faster hardware interfaces (VHBR).

---

5. Comparative Analysis of Cryptographic Transitions

Comparing the migration timelines reveals significant regional disparities.

Algorithm Transition	Estonia (EstEID)	USA (PIV)	EU (eID/Passport)	Japan (JPKI)	Lag (Japan vs Leaders)
RSA 1024 → 2048	2011	2013 (FIPS 201-2)	~2010-2012	2016 (My Number Card)	3-5 Years
RSA → ECC	2018 (Gen 3)	2013 (Option)	2021 (Mandatory)	~2030 (Next-Gen)	5-8 Years
PQC (Post-Quantum)	Watching	Piloting	2024 (German Pilot)	TBD	Unknown

5.1. The "Japan Lag" Hypothesis

There is a concern that Japan is "10 years behind." The data suggests a 5-8 year lag in adopting the latest cryptographic primitives (ECC), rather than a full decade.

• Factor 1: Conservative Lifecycle: JPKI adheres strictly to ISO 7816 with RSA 2048, favoring stability over agility.
• Factor 2: Mobile Integration: While Estonia and the US moved to ECC to support mobile/contactless performance, Japan's "Smartphone JPKI" (2023) currently emulates the existing RSA 2048 card structure.

Official Roadmap vs. Operational Reality According to the Digital Priority Plan 2024, the government targets the Next-Gen Card introduction in 2026 and aims for a cryptographic transition aligned with the 2030 mass expiration. However, the physical reality of card issuance suggests a longer migration tail.

• The Government Goal: RSA 2048 disallowed by Jan 1, 2031.
• The Operational Reality:
  • Chip Constraint: Existing RSA-based cards cannot generate ECC keys. When a user renews their certificate (every 5 years) on an old card, the new certificate must still be RSA.
  • The "Hybrid Era" (2030-2040): Even if ECC cards become standard in 2030, previously issued RSA cards will remain valid for up to 10 years (or 5 years for certificates). Unless a mandatory recall is enforced, RSA support will be required in the field until at least 2035 (certificate expiry) or 2040 (card expiry).

Conclusion: Japan faces a decade-long "Hybrid Era." The civ library must support both RSA and ECC simultaneously for a very long time. We cannot simply "switch off" RSA support in 2031; robust dual-stack support is the critical architectural requirement for the next 15 years.

6. Regional Deep Dive & Recommendations for Civ

6.1. Europe (The "PACE" Standard)

Europe is converging on ICAO 9303 + PACE.

• Observation: The "National ID" and "Passport" technologies are merging.
• Civ Action: Prioritize PACE implementation in Icao9303Controller. This single feature unlocks support for almost all modern European ID cards.

6.2. Japan (JPKI)

Japan remains on RSA 2048 (ISO 7816 based).

• Observation: Solid, proven, but lagging behind the global ECC trend.
• Civ Action: Monitor JPKI spec revisions for ECC adoption. Current JpkiController is stable but should prepare for the "Next Gen" card (c. 2026).

6.3. Southeast Asia (The "Fragmentation" Zone)

Thailand and Malaysia use distinct, older architectures.

• Thailand: Simple structure. High value due to tourism/expat use cases.
• Malaysia: Complex (multi-app).
• Civ Action: Implement dedicated ThaiController (Simple) and MyKadController. These cannot share logic with ICAO/JPKI drivers.

6.4. USA (PIV)

PIV is a robust standard (FIPS 201).

• Observation: Widely used in government/defense.
• Civ Action: PivController is a high-value addition for enterprise/gov use cases.

6.5. USA (PIV) - Post-Quantum Transition (PQC4PIV)

The US is leading the PQC standardization for ID cards. While the algorithms are finalized (FIPS 204/205), the card specification is currently in Draft (NIST SP 800-73-5 Revision).

• Primary Algorithms:
  • ML-DSA (Module-Lattice-Based Digital Signature): Derived from CRYSTALS-Dilithium. Primary candidate for document signing and authentication due to balanced speed and key size.
  • SLH-DSA (Stateless Hash-Based Digital Signature): Derived from SPHINCS+. Secondary candidate, offering conservative security based on hash functions but with larger signatures.
• Draft Specifications (SP 800-73-5):
  • New Key Types: Defining new Algorithm Identifiers (OIDs) for ML-DSA-44/65/87.
  • Hybrid Credentials: A transition strategy allowing a single PIV card to host both classical (RSA/ECC) and PQC certificates to ensure backward compatibility.
  • Hardware Challenges:
    • Key Size: PQC keys (esp. Dilithium) are significantly larger than RSA/ECC.
    • Transmission: Requires robust support for Extended Length APDUs (up to 32KB/64KB) to read certificates and signatures efficiently. While JPKI already utilizes Extended Length for photo and certificate retrieval, PQC payloads may push these limits further.
    • Bus Speed: Pushing for VHBR (Very High Bit Rate) protocols to reduce transaction times for large data payloads.

Civ Implementation Strategy: Verify and harden Extended Length APDU support. Although civ's APDU builder seemingly handles variable lengths, PQC's data sizes might exceed standard buffer limits or require specific T=1 protocol optimizations (chaining) that need explicit testing against next-gen card specs.

7. Summary Timeline

Era	Crypto	Representative Schemes	Civ Support
2000-2010	RSA 1024, 3DES, SHA-1	EstEID (Gen1), MyKad (Gen1), ICAO BAC, Juki-Card	Legacy (BAC supported)
2011-2015	RSA 2048, SHA-256	EstEID (Gen2), US PIV (RSA)	Good
2016-2020	RSA 2048, SHA-256	JPKI (Current), ICAO PACE (Early)	Good (JPKI)
2021-2025	ECC P-256/384, AES	EU eID (PACE), EstEID (Gen3), US PIV (ECC)	Good (PACE)
2026+	PQC (Kyber), ECC Hybrid	German PQC Pilot, US PIV (PQC Pilot)	Future Research