Publications

Peer-reviewed research publications from the Veralix Foundation research program. All methodologies are published under open access principles.

Blockchain Forensics2026

Sliced-Wasserstein Graph Kernels for Blockchain Anti-Money Laundering: A Daubert-Compliant Framework

Veralix Research TeamJournal of Digital Forensics, Security and Law

We present a novel approach to blockchain transaction graph analysis using Sliced-Wasserstein distances combined with Weisfeiler-Lehman graph kernels. Our method achieves 94.7% detection accuracy with a false positive rate of 0.3%, meeting the Daubert standard for scientific reliability in legal proceedings.

Cryptography2026

Five-Layer HMAC-SHA-256 Audit Chains: Cryptographic Chain-of-Custody for Digital Evidence

Veralix Research TeamIEEE Transactions on Information Forensics and Security

This paper introduces a five-layer parallel hash-chain architecture for maintaining tamper-evident audit trails in digital forensic workflows. We prove that any single-bit modification to evidence or metadata is detectable with probability 1 - 2^(-256) per layer.

AI & Law2026

Human-in-the-Loop AI for Legal Admissibility: Bridging Machine Learning and the Federal Rules of Evidence

Veralix Research TeamArtificial Intelligence and Law

We propose a framework for integrating AI analytical capabilities into legal evidence workflows while maintaining strict compliance with FRE 702 and Daubert criteria. Our Human-in-the-Loop architecture ensures AI outputs serve as analytical aids requiring mandatory human expert validation.

Security Architecture2026

Ghost Mode: Zero-Persistence Architecture for Sensitive Evidence Processing

Veralix Research TeamACM Conference on Computer and Communications Security

We present Ghost Mode, a zero-persistence computing architecture that processes sensitive evidence entirely in volatile memory with cryptographic destruction guarantees. We demonstrate that no residual data persists beyond session termination through formal verification of memory erasure protocols.

Blockchain Forensics2025

Cross-Chain Forensic Tracing: Entity Resolution Across Heterogeneous Blockchain Networks

Veralix Research TeamFinancial Cryptography and Data Security

This work addresses the challenge of tracing financial flows across multiple blockchain networks with different consensus mechanisms and address formats. Our layered approach achieves 89.2% entity resolution accuracy across Bitcoin, Ethereum, and EVM-compatible chains.

Legal Technology2025

Admissibility Checking Engine (ACE): Automated Assessment of Digital Evidence Against Legal Standards

Veralix Research TeamInternational Conference on AI and Law

ACE provides automated pre-screening of digital evidence against jurisdiction-specific admissibility requirements including authentication, best evidence rule, hearsay exceptions, and scientific reliability standards. Field testing with 200 evidence items showed 96.1% agreement with expert attorney assessments.