RESEARCH
Post-quantum research doubles and moves from algorithms to deployment
A week dominated by cryptography: post-quantum papers rise to 35, more than double the average, with hardware accelerators and parameter choices for ML-KEM and ML-DSA. Side channels keep targeting post-quantum implementations, cryptography for blockchains grows and AI security falls in volume. Attention emerges for compilers and sensors compromised before use.
This week's trends
Post-Quantum Cryptography · Rising
35 papers this week · average of the previous 4 weeks: 15.2
Post-quantum papers rise to 35, more than double the 15.2 average. The focus moves from algorithms to deployment: Adams Bridge is an industry-grade hardware accelerator combining ML-KEM and ML-DSA, Kaiburr discusses ML-KEM parameters at much higher security, and another paper measures archive certificates for ML-DSA-87 signatures on blockchains. Lattice-based anonymous tokens and identity-based encryption complete the picture. For insurers, the post-quantum migration is entering the phase of concrete product choices.
Papers cited:
- Adams Bridge (V2.0) Accelerator: Bridging the Post-Quantum Transition
- Kaiburr: ML-KEM at (much) higher security
- Archive Certificates for ML-DSA-87 Authorizations from an Exact Bounded Relation
- Atlantis: Lattice-based Anonymous Tokens with Private Metadata Bit
- An efficient Failure Probability Estimator for LWE-based encryption, and application to IBE
Hardware & Side Channels · Steady
8 papers this week · average of the previous 4 weeks: 8.2
The count is stable (8 versus 8.2) and, as last week, side channels target post-quantum schemes. Correlation power analysis recovers portions of the SHA-2-based SLH-DSA seed on a Cortex-M4; another paper exploits the structure of syndrome decoding to rebuild the SDitH secret from imperfect side-channel data. A third models multi-party computation with a single enclave. The message for the market is unchanged: the post-quantum robustness of devices depends on the implementation.
Papers cited:
- Correlation Power Analysis on SHA-2-Based SLH-DSA
- Syndrome Decoding out of the Head: Structure-Aided Side-Channel Witness Recovery on SDitH
- More from Less: Composable General Multi-Party Computation with Global Public Verifiability from a Single Enclave Only
Cryptography · Rising
70 papers this week · average of the previous 4 weeks: 43.5
Cryptography rises to 70 papers against a 43.5 average, and a visible share concerns public ledgers. One paper proves key provenance in hierarchical deterministic wallets without revealing the seed, a SoK examines cryptographic erasure on blockchains against GDPR Article 17, and another assesses homomorphic encryption on blockchains for medical consultations. A fourth formalises the post-quantum signature transition for proof-of-work blockchains. For insurers, compliance and key custody remain the critical points.
Papers cited:
- Provenance Proofs: Linkable Zero-Knowledge Derivation Relations for Hierarchical Deterministic Wallets
- SoK: Cryptographic Erasure on Public Ledgers: Application-Layer Architectures, Key-Lifecycle Adversaries, and GDPR Art. 17 Sufficiency
- Can Fully Homomorphic Encryption on Public Blockchains Protect Sensitive Medical Consultations? A Security and Privacy Assessment
- Formalizing PQC Signature Transition: Case of PoW Blockchain Against On-Spend Attack
AI & LLM Security · Cooling
53 papers this week · average of the previous 4 weeks: 116.8
AI security drops to 53 papers against a 116.8 average. In a single week the fall should be read with caution and does not signal lost interest: among the strongest papers, SimBreak defeats TraintoDefend, the only defense believed effective against cryptanalytic model extraction, and a technical report proposes three practical defenses against data-poisoning backdoors in face recognition pipelines. Work on agents and MCP tools remains present in the curated edition.
Papers cited:
- Breaking \mathsf{TraintoDefend} via Adaptive Neuron Clustering in Cryptanalytic Model Extraction
- A Bag of Tricks for Defending Deep Learning-based Face Recognition Pipelines against Backdoor Attacks
Software Supply Chain · Emerging
1 papers this week · average of the previous 4 weeks: 2.8
Two strong papers shift attention to components trusted before execution. One shows that known compiler bugs make correct zkVMs produce valid proofs of programs that differ from the source, and that innocuous-looking patches can trigger them. The other, SenTrojan, implants analog Trojans in sensors before deployment. In both cases code review and ordinary testing fall short. For insurers, the provenance of toolchains and components becomes an underwriting question, with possible accumulation.
Papers cited:
- Proving the Wrong Program: Miscompilations as a Threat to zkVM Guarantees
- SenTrojan: Revealing Analog Hardware Trojan Attacks on Sensors
How this digest is made
New papers are collected every day and selected every week from arXiv, the IACR ePrint archive and the main security conferences and journals (IEEE S&P, USENIX Security, ACM CCS, NDSS, CRYPTO, PoPETs, IEEE TIFS and others). A score based on source, topic and early signals picks the candidates; an AI model assesses them and writes the summaries. Summaries are based on the abstracts: the original paper is the reference for details.
New papers assessed for this edition: 321.