PIR Wednesday — August 5 2026 Weekly Update

What Was Done

New Files Created

1. SimplePIR and DoublePIR (~9 KB) — Full deep dive on the foundational USENIX Security 2023 schemes. Covers: core idea (LWE-based "square-root" PIR with Regev encryption), setup and online query mechanics, SimplePIR vs DoublePIR comparison (throughput, hint size, communication), extensions (long records, batch queries), the full SimplePIR family evolution table (YPIR, VeriSimplePIR, ThorPIR, iSimplePIR, RotPIR, ZipPIR, FlashPIR), code reproduction instructions, and limitations/open problems.

2. PIR Implementations and Libraries (~14 KB) — Comprehensive tracker of every known PIR implementation. Organized by era: Foundational (SealPIR, XPIR, FastPIR, Spiral), SimplePIR Family (SimplePIR, YPIR, ThorPIR, WhisPIR, VeriSimplePIR, Respire), Hintless/Stateless (Google hintless_pir, FrodoPIR, OnionPIRv2), 2025–2026 Frontier (keyword PIR: ChalametPIR/KPIR/Conflux/PIRacle; hardware-accelerated: GPIR/SmartPIR/IM-PIR; communication-optimized: FlashPIR/ZipPIR/VIA/InsPIRe/Pirouette/TensorPIR; preprocessing-optimized: RotPIR/iSimplePIR/PIRCOR), multi-server (mpc4j, hybridpir, Piano, Tiptoe), and survey resources.

3. PIR Use Cases and Deployments (~16 KB) — Full deployment analysis covering 7 use cases: CT Log Auditing (YPIR 48× cost reduction), Ethereum Private RPC Reads (Kohaku, sharded PIR design, RisePIR, PSE project, TEE+ORAM alternative), Private Web Search (Tiptoe SOSP 2023), Apple Device Features (iOS 18 Live Caller ID, production), Password Breach Checking (iSimplePIR), Private Contact Discovery (Signal, Respire), Private DNS. Includes use case maturity matrix and 4 deployment architecture patterns.

Updated Files

4. Overview - Private Information Retrieval — Expanded from ~8.4 KB to ~17.5 KB (108% increase). Status upgraded from developing to mature. Major additions: - Complete rewrite of historical context with 2024–2026 papers (YPIR, ThorPIR, WhisPIR, VeriSimplePIR, Respire, FlashPIR, ZipPIR, RotPIR, iSimplePIR, OnionPIRv2, PIRCOR, VIA, Improved TensorPIR, InsPIRe, Pirouette, Conflux, PIRacle, GPIR, SmartPIR, KPIR, IM-PIR). - New implementations table with 20 entries (up from 9). - Expanded use cases with Ethereum/Kohaku integration details. - New "Research Frontier (August 2026)" section with 7 active open problems. - Updated evidence-at-a-glance table with 21 schemes (up from 11). - Added Tiptoe, Apple deployment, password breach checking as use cases.

Key Research Findings (August 2026)

The Field Has Diversified Explosively

The PIR research landscape in 2025–2026 is no longer just about raw throughput. The frontier has split into seven active subproblems:

1. Incremental / dynamic databases — iSimplePIR (entry-level updates), Conflux (keyword PIR for dynamic data), PIRacle (key-value stores). The static-DB assumption that limited PIR's practicality is being dismantled.

2. Keyword PIR — KPIR (USENIX Sec 2025), Conflux (HPCA 2026), ChalametPIR. Practical systems need keyword lookup, not just index-based retrieval. KPIR provides a generic transformation from keyword to index PIR via sparse key-value stores.

3. Verifiability — VeriSimplePIR (USENIX Sec 2024, 27 citations), "Verifiable PIR with Small Client Storage" (S&P 2026). Detecting malicious servers that return wrong results. The S&P 2026 follow-up reduces client state from >1 GiB by orders of magnitude and introduces covert vPIR for stateless clients.

4. Hardware acceleration — GPIR (GPU, 305×), SmartPIR (CSDs, MICRO 2025), IM-PIR (processing-in-memory). PIR's memory-bound nature makes it ideal for near-data processing architectures.

5. Communication reduction — FlashPIR (2.6× over WhisPIR), ZipPIR (no client storage), VIA (S&P 2026), Pirouette (transciphering alternative), InsPIRe, Improved TensorPIR (45.5% reduction).

6. Preprocessing elimination — YPIR (silent preprocessing), WhisPIR (stateless), ZipPIR (no client storage at all). Removing the hint download barrier.

7. Ethereum integration — The first major platform-level deployment target. Sharded PIR design (ethresear.ch), Kohaku privacy toolkit (Vitalik endorsed), PSE PIR project, RisePIR for key-value workloads.

Key Papers Discovered This Session

Ethereum PIR — The Deployment Signal

The strongest real-world deployment signal for PIR comes from Ethereum:

• Vitalik Buterin's privacy roadmap (May 2026): Explicitly calls for PIR + ORAM for wallet privacy. Kohaku toolkit introduced as the implementation vehicle.
• Sharded PIR Design (ethresear.ch, Ali Atiia & Keewoo Lee): Comprehensive architecture for pairing Ethereum data slices with optimal PIR schemes. Proposes decoy queries across all shards for uniform privacy.
• RisePIR: Keyword PIR for key-value stores using Segmented Cuckoo Filters. Maps directly to (address → balance) lookups. Single-threaded prototype achieves 1,500–24,000 ops/s for 32-byte values — squarely in the regime needed for eth_getBalance.
• PSE Project: Active Ethereum Foundation team building PIR for Ethereum data (pse.dev/blog/ethereum-privacy-pir, June 2025).

Apple — The Existence Proof

Apple's iOS 18 (2024) ships HE+PIR for Live Caller ID, logo lookup, and content filtering. This is the largest-scale production deployment of PIR-class techniques in consumer hardware. It proves the approach works at consumer scale on mobile devices, though Apple's implementation is proprietary and not independently auditable.