"Genomic sequences have cryptographic secrets too"

To our knowledge, DNAcrypt-AI is the first system to leverage the human genome and AI together to generate, encrypt, and decrypt digital information in a reproducible, high-entropy manner, a milestone not just for biology and AI, but for the rapidly expanding digital world of today and tomorrow.

— Marvin De los Santos (Project lead, DNAcrypt-AI)

Research Overview

Cryptography using the genome and AI is an emerging and interesting research direction that explores how the structural complexity, scale, and organization of the human genome can be formally repurposed for digital security systems. Our reserch serves as a reference and successful implementation of this field. With DNAcrypt-AI, we demonstrate how genome-derived vocabularies, sequence reconstruction, and AI-driven interpretation can be integrated into a reproducible and efficient encryption-decryption pipeline.

Scientific Motivation

Modern digital systems rely on cryptographic methods that are predominantly mathematical and rule-based. While highly effective, these systems operate within relatively constrained and well-studied design spaces. In contrast, biological systems, particularly the human genome, exhibit massive scale, non-uniform structure, and contextual complexity that remain computationally challenging to model exhaustively.

Although the human genome has been extensively decoded, its full combinatorial and structural landscape remains far from trivial to exploit computationally. Our research investigates how such complexity can be leveraged for non-biological interpretation and as a foundation for cryptographic entropy and controlled information concealment.

Research Objectives

We are interested in the following goals:

• Formally define how genome assemblies can function as cryptographic coordinate spaces

• Establish genome-derived vocabularies as high-entropy substrates for encryption

• Demonstrate deterministic yet non-reversible transformations using biological data

• Justify the role of AI as a necessary interpreter between genomic sequences and digital symbols

• Evaluate reproducibility and traceability using public reference genomes

Preliminary Foundation

DNAcrypt-AI operationalizes our research by implementing the methodical frameworks that we pioneered for human genome-based cryptography in the era of AI. It integrates genome vocabulary generation, sequence reconstruction, and AI-based decoding into a single system capable of generating, encrypting, and decrypting passwords and cryptographic keys. Although, the successful deployment of DNAcrypt-AI marks the beginning of new era of bioencrpytion, it is important to not that DNAcrypt-AI is not positioned as a replacement for established cryptographic standards just yet. However, it is a proof-of-concept that exploring how biological information systems and artificial intelligence can inform new security paradigms is within reach.

Learn more about the architecture and process flow of DNAcrypt-AI.