EviDNA DNA Cryptography | Jacques Gascuel Memory

EviDNA DNA cryptography — executive summary

This complementary thesis documents the Freemindtronic trajectory in the cryptographic universe mobilizing the expression “DNA” in the procedural and architectural sense — non-molecular by default: EviDNA (human profile, 2024), ADN Digital, cryptographic genome and industrialization CryptPeer/EviSKMS (2026).

It establishes documentary comparisons with the state of the art: classic digital trust mechanisms (FIDO, PKI, Zero Trust, HSM/TPM), academic genomic data encryption (PROMISE, Varlock), and institutional approach CNRS 2026 (synthetic DNA, OTP/Vernam). He does not claim any authorship on the third-party works; It specifies distinct technical objects. Canonical definition EviDNA: §1.11.

The publication respects the registers A (public), B (confidential) and C (IP): two international patents granted are publicly cited (WO/2018/154258 — segmented key; WO/2017/129887 — access control); no records enabling the reproduction of EviDNA, genome, Gen2 or advanced runtime mechanisms (C registry).

For the interdisciplinary framework linking predictive AI, cybersecurity, and cyber-physical trust, see EviSKMS reference memory.

EviDNA DNA cryptography — Relation to the “predictive intelligence architectures — EviSKMS”</h2 thesis>

The two dissertations are complementary: the first sets the broad scientific framework; The second deepens the cryptographic trajectory and state-of-the-art comparisons without diluting the debate on artificial general intelligence.

1. Cryptographic genome, EviDNA and industrial trajectory

In the context of this thesis, the expression “cryptographic genome” does not refer to biological DNA, nor to a direct exploitation of biometric data, nor to a form of DNA computing. Nor does it refer to a new fundamental cryptographic building block intended to replace existing standards, encryption algorithms, signature mechanisms, PKIs, HSMs, TPMs or digital identity repositories.

It refers to a digital trust architecture approach aimed at organizing, over time, evidence, contexts, policies, states of trust, and local and online verification mechanisms around a continuity of trust. This does not prescribe a single encryption algorithm: it is agnostic with respect to cryptographic bricks — symmetric (including OTP / single-use masks), asymmetric, post-quantum (PQC), etc. — in accordance with the governance policy. It should be understood as a structuring, governance and verifiability, and not as a substitute for existing cryptographic standards.

A first generation of this approach is already industrialized in CryptPeer via EviSKMS. It materializes, at an operational level, a segmented, locally verifiable, policy-driven, and runtime-oriented trust. This Gen1 is a return to industrialization: it demonstrates that an identity, a session, an execution context or a trusted object can be treated not as a simple static identifier, but as a controlled, reassessable and governable trust structure.

Jalon EviDNA — three-step timeline (registry A).

Synthetic chronology (text schema, register A).

2017 ──► QR chiffré + NFC M24LR (commercial, sans couche ADN)
           │
2022 ────► Eurosatory — seed / EviDNA project (R& D)
           │
2022-24 ─► ST25 64K +EviDNA Development
           │
2024 ────► Eurosatory Lab — DataShielder Defense NFC HSM (industrialisé)
           │
2024-26 ─► Digital DNA + giscryptographique name
           │
2026 ────► Eurosatory — CryptPeer/EviSKMS industrialisé · TPM/vTPM

Defense / EviDNA detail: §1.11 · Product Proof§1.10. Digital DNA / CryptPeer 2026: §1.7.

To preserve scientific rigor, the qualification of industrialized Gen1 must remain attached to observable elements: code, frozen contracts, tests, runtime flows, implementation logs, technical documentation or product integration. Unpublished implementation details are not set out in this supplementary brief.

1.1. Non-sensitive level of evidence and Gen1</h4 industrialization perimeter> This subsection is part of the same methodological logic: it does not aim to impose recognition by personal authority, but to link an inventor’s intuition to verifiable, non-sensitive and observable elements. The weak and strong signals identified in the field serve here as raw material for a cautious scientific formalization, without enabling disclosure of internal mechanisms.

This thesis does not seek to publish the internal mechanisms of the cryptographic genome. It establishes its scientific and industrial positioning: a segmented, local, temporal and governable digital trust architecture, whose Gen1 and Gen2 are industrialized in CryptPeer via EviSKMS.

In order to avoid any enabling technical disclosure, the evidence mentioned below is formulated at a non-sensitive level. They indicate the scope of industrialization without exposing the detailed mechanisms, internal structures, operational formats, verification sequences or transition rules.

Patented, publishable parentage. The principle of segmented key and conditional reconstitution of trust can be publicly cited under the international patent WO/2018/154258 (FR3063365 B1, EP3586258, US20210136579, CN110402440, JP2020508533, KR1020190120317). This foundation covers segmentation, physical proximity, token, ephemeral volatile memory, segment governance and a variant of the invention — the scrambling module of authentication data — without allowing the disclosure of post-patent extensions not yet registered (genome, detailed EviDNA, advanced runtime).

1.1.1. Jamming module — public variant of patent (WO/2018/154258)

The granted international patent WO/2018/154258 (FR3063365 B1, EP3586258B1) describes, in addition to the segmented key, a variant of the invention relating to a scrambling module authentication data. This mechanism is freely accessible in the public description of the title: when typing on an untrusted channel (keyboard, interface, clipboard), additional characters are inserted at predetermined positions known to the legitimate user, who removes them before transmission. The documented objective is to reduce the exposure of the real secret in the face of a keylogger or any direct observation of the input surface.

Cryptographic positioning (ledger A). This module is not an OTP/Vernam schema: it protects the transient representation of the secret at the time of input, not the content of an encrypted message.

Limits and C.</strong registry> Any auto-extension, runtime generalization, or correlation with EviDNA, cryptographic genome, or EviSKMS falls under the C registry as long as no additional repositories are secured. This paragraph is limited to the variant of the issued title.

Classification legend: A = possible audience in the memory · B = confidential (private file, audit under NDA) · C = reserved IP (before filing or validation by patent advisors).

This matrix does not purport to be a complete technical publication. It establishes a level of maturity that can be read by the scientific reader: the Gen1 and the Gen2 are industrialized in CryptPeer, anchored on an international patent issued for segmentation; the detailed mechanisms of Gen2 fall under the C register.

Full scientific recognition of this approach will require additional publications, intellectual property filings when necessary, as well as comparative evaluations documenting its contributions to traditional authentication, passwordless, PKI, access control and runtime trust mechanisms.

1.2. Towards controlled scientific recognition: evidence, comparisons and publication after PI</h4 securitization> The full scientific recognition of this approach presupposes a complementary step, carried out after securing intellectual property when necessary. This stage will have to articulate three levels: non-sensitive evidence of industrialization, structured comparisons with the state of the art and controlled publication. A first appendix of non-sensitive evidence, resulting from a local analysis of the EviSKMS-CryptPeer repository, now makes it possible to document this first level without exposing the internal mechanisms protected.

Non-sensitive evidence will be able to document the existence of operational implementation without disclosing the protected internal mechanisms. They may include product scope, functional architecture, maturity levels, usage scenarios, general flows, test categories, trust policies, execution logs, and validation criteria.

Comparisons will have to situate the Freemindtronic approach in relation to the existing mechanisms of authentication, passwordless, PKI, HSM, TPM, Zero Trust, WebAuthn/FIDO externally, machine identity, IoT and runtime trust. The objective will not be to replace them with affirmation, but to show where the genomics approach to digital trust brings a different layer: segmentation, local verification, temporal continuity, contextual governance and reassessment of the level of trust. A first comparative document matrix is proposed in §1.4.

The controlled publication can then take the form of a position paper, a scientific white paper, an evaluation report or a documented demonstrator. It should remain non-enabling until intellectual property protections are finalized, while providing sufficient elements to allow scientific discussion: problem addressed, hypotheses, scope, comparison, limitations, use cases and evaluation protocol.

Publication doctrine (register A). This thesis deliberately adopts a controlled publication logic: it documents scientific subject-matter, prior art, state-of-the-art comparisons and evidence of industrialization observable, without disclosing the internal mechanisms that may be the subject of complementary patent filings. This applies in particular to the advanced implementation in CryptPeer/EviSKMS, where only functional effects, architecture principles, and non-sensitive elements are exposed. The rules of derivation, transition, genomic correlation, internal formats and operating parameters remain in the B or C register. Detail: §1.12.

This trajectory makes it possible to clearly distinguish three registers: what is already industrialized, what can be made public without risk to intellectual property, and what must remain reserved for deposits, confidential annexes or evaluations under confidentiality agreements. It thus avoids two opposing pitfalls: an unproven assertion of innovation, or a premature disclosure of protected technical mechanisms.

The Gen2 is implemented in CryptPeer via EviSKMS. It extends the Gen1 trajectory towards an evolving, contextual, memory and verifiable digital identity over time. The detailed technical mechanisms fall under the C registry and are not disclosed in this supplementary submission.

The emergence of predictive artificial intelligence makes this development particularly important. Attacks are no longer just about isolated passwords or certificates. They can target identity continuities: progressive spoofing, deepfakes, session compromise, hijacking of AI agents, cloning of connected objects, context alteration, memory poisoning or behavioral manipulation.

Faced with these risks, one-time authentication becomes insufficient. A future identity architecture will need to verify not only what an entity knows, owns, or is, but also the context in which it operates, the consistency of its interactions, the governance of its rights, the continuity of its evidence, and the reassessment of its level of trust over time.

The cryptographic genome thus constitutes a two-stage trajectory: a Gen1 and a Gen2 industrialized in CryptPeer via EviSKMS. Gen1 embodies segmented, local and runtime-governed trust; Gen2 extends this approach to an evolving and contextual identity. Gen2 technical details are protected when they are likely to fall under additional intellectual property protections.

This approach should be thought of as distinct from the FIDO/Passkeys mechanisms, which Freemindtronic does not use as a foundation of trust. It can be situated in relation to existing repositories—NIST SP 800-63-4, Zero Trust, ETSI EN 303 645, Cyber Resilience Act, and, for external comparison, WebAuthn/FIDO—but not limited to or dependent on it.

Freemindtronic is also developing its own passwordless approach, based on EviSKMS and the Gen2 evolution. In order to preserve current or future intellectual property protections, this brief does not disclose the detailed technical mechanisms.

The public positioning can nevertheless be formulated as follows: this digital trusted genomic technology aims for a segmented, local, temporal and verifiable approach to identity and authentication. It is intended to apply to many contexts where it becomes necessary to establish, maintain or reassess a trusted identity: humans, connected objects, software agents, digital services, cyber-physical environments, critical access, secure exchanges and runtime continuity.

Its interest lies in the fact that it no longer considers identity as a simple one-off authentication event, but as a continuity of trust that is evolving, governable and verifiable over time. This orientation becomes especially important in contexts where traditional passwordless mechanisms and traditional authentication are becoming insufficient in the face of predictive AI, autonomous agents, synthetic identities, session compromises, and behavioral attacks.

This perspective is in line with the general axis of this thesis: predictive AI transforms the conditions of trust. The more systems become capable of anticipating, acting and adapting, the more identity itself must become reassessable, memorial, contextual, verifiable and governable over time.

1.3. EviSKMS-CryptPeer</h4 industrialization proof-of-the-mill summary> A synthesis of evidence of industrialization was established from a local analysis of the EviSKMS-CryptPeer repository. It does not reproduce any source code, pseudo-code, operational format, verification sequence, transition rule or repeatable mechanism. Its goal is to provide the scientific reader with proof of existence and maturity, without enabling disclosure.

This appendix confirms that CryptPeer is an integration and operational governance layer aligned with EviSKMS. It documents, at a high level, the existence of a trusted runtime, Runtime Integrity controls, DRT continuity, sovereign runtime certificate (RSCC), fail-closed policies, anti-replay guardrails, chained logs, cryptographic governance, compound proofs, frozen sovereign passwordless mode V1.1, DDNA Gen1 foundation, automated security testing campaign, and sovereign deployment artifacts.

Filiation brevete. The observable industrialization is in line with the international patent Segmented Key Authentication System (WO/2018/154258, FR3063365 B1). This title allows for the public disclosure, without weakening the residual IP, of the principles of segmented key, physical proximity, conditional reconstruction, protection of authentication data and the variant of the jamming module (§1.1.1) — the foundation on which EviSKMS and CryptPeer have been industrialized. The extensions genomic Gen2, the engine DRT complete, the convergence multi-criteria advanced, and non-patented internal mechanisms remain outside the public perimeter.

The scientific value of this synthesis does not lie in the disclosure of internal mechanisms, but in the methodological distinction between three registers:

Disclosure perimeters (text schema).

                    ┌─────────────────────────────────────┐
                    │ C — Reserved PI │
                    │ Gen2, Continuity Engine (internal), runtime extensions (internal) │
                    │ passwordless, genome transitions │
                    │  ┌───────────────────────────────┐  │
                    │ │ B — Confidential / NDA │ │
                    │  │ runbooks, red team, code privé│  │
                    │  │ ┌─────────────────────────┐   │  │
                    │ │ │ A — Public (memory) │ │ │
                    │  │ │ brevet, fail-closed,    │   │  │
                    │ │ │ │ High-level events │ │ │
                    │  │ └─────────────────────────┘   │  │
                    │  └───────────────────────────────┘  │
                    └─────────────────────────────────────┘

EviSKMS–CryptPeer Stacking (Text Schema, A Register).

Applications / opérateur
        │
        ▼
CryptPeer — governance, integration, sovereign deployment
        │
        ▼
EviSKMS runtime ──┬── Runtime Integrity (RI) / RSCC
                  ├── DRT (continuity of trust)
                  ├── DDNA Gen1 (empreintes normalisées)
                  ├── Passwordless V1.1 (sovereign-local)
                  └── Fail-closed · Anti-Replay · chained newspapers
        │
        ▼
Hardware Anchor: TPM / vTPM (2026) — segments, policies

Directly usable public evidence (Registry A): EviSKMS–CryptPeer architecture; software-sovereign-first ecosystem gel; Runtime Integrity and RSCC as posture artifacts; built-in DRT continuity; multi-surface anti-replay; Logs with separate rolls. passwordless V1.1 qualified sovereign-local; DDNA Gen1 by standardized impressions; security test campaign; Filiation patent WO2018154258.

Do not publish: code, pseudocode, canonical payloads, check sequences, transition rules, red team fixtures, secondary segment details, advanced multi-criteria composition, Gen2.

This separation supports the credibility of the brief — and the associated industry communications — without turning the public document into a technical reproduction record. It establishes that the Gen1 of the cryptographic genome has a double anchor: an international patent granted on segmentation, and industrialization observable in CryptPeer via EviSKMS.

The exact scope of this evidence is deliberately limited: it does not constitute independent scientific validation or peer review. However, it constitutes a sufficient documentary basis for a controlled publication, a white paper, an evaluation report or a client file, after securing the patentable elements that have not yet been filed. The limits and conditions of falsifiability of the brief specify what this proof does not establish.

1.4. Structured comparison — digital trust and identity

This subsection responds to the need, formulated in §1.2, of an explicit comparison with the state of the art in terms of digital trust. It is not a quantified performance benchmark, nor a third-party audit, but a documentary positioning at a non-enabling level.

Scope compared. The following are compared, at a high level: WebAuthn / FIDO / Passkeys (external comparison — Freemindtronic does not use FIDO as a trust base), PKI / X.509, Zero Trust (NIST framework), HSM / TPM, OAuth / Federated OIDC, and EviSKMS Gen1 / CryptPeer as documented in the A</strong register> in this supplementary submission and the Appendix C.

Qualitative rating: Low · Medium · Strong · Very strong · N/A (not applicable to the perimeter).

Methodological reading. This table does not classify EviSKMS as “superior” on all axes. It shows a difference in function:

• FIDO/OAuth/PKI excel at interoperability, standardization and large-scale one-time authentication
• Zero Trust provides a framework for governance and policies, but is not a local sovereign trust runtime on its own.
• HSM / TPM reinforce the material anchor, often in addition to other layers.
• EviSKMS Gen1 aims for an layer additive: trust segmented, continuous over time, verifiable locally and governed to the runtime, as an extension of the segmented key patent — at the cost of less immediate interoperability and independent scientific validation still to be conducted.

What the comparison does not establish. It does not demonstrate the operational superiority of EviSKMS over FIDO or PKI in all contexts. It does not replace comparative numerical trials, published red team campaigns or certification. It situates the Freemindtronic positioning for a structured scientific and industrial discussion.

1.5. Cryptographic genome vs. point identity (time T)

Verification of the distinction. Recent institutional work on synthetic DNA and OTP (CNRS communication April 2026, HAL hal-05560338) describe a protocol where two correspondents have identical copies of synthetic DNA sequences, then just before a communication select and sequence fragments to produce a common binary key at time T — key distribution logic synchronized to an event, not a identity architecture evolving over time. The classic authentication mechanisms (password, certificate, WebAuthn, point biometrics) obey the same functional structure: prove “it’s me” at the moment T, then grant or deny access.

The Freemindtronic cryptographic genome is part of a different technical object: a digital trust architecture that organizes, over time, proofs, contexts, policies, runtime states, normalized fingerprints (DDNA Gen1), session continuity, fail-closed reevaluation and — in Gen2 — contextual identity, Memory and governable. This is not a marketing metaphor for molecular DNA: the expression refers to a procedural structuring of trust (segments, inheritances, dependencies, traceability), publicly formalized in this thesis and initiated by EviDNA (2024) then ADN Digital (2026).

Time horizon: time T vs continuity (text diagram).

 punctual auth / CNRS OTP (time T) Cryptographic genome (continuity)
────────────────────────────────────          ────────────────────────────────────

    T₀ T₀ T₁ T₂ Tn
     │                                                │         │         │         │
 [Proof] ──► Granted or refused?       [Confidence inValuable ─────────────►]
     │                                                │
     ✕ (end of event) fail-closed · DDNA · DRT · segments

Synthèse. This precise distinction between distinct technical objects: the CNRS mobilizes synthetic DNA to a single scheme (OTP/Vernam at a given time); The Freemindtronic trajectory can also produce OTP keys, but in a broader architecture — segmented and continuous trust over time, with interchangeable mechanisms. The Freemindtronic Public Disclosures (2018–2026), the online submission (freemindtronic.com) and the patent WO/2018/154258 are elements of documented prior art on this trajectory. For the CNRS approach as publicly formulated, see §1.6.

1.6. Documentary synthesis — CNRS DNA cryptography (external reference, register A)

Status. This subsection does not claim any authorship on CNRS work. It faithfully transcribes, for documentary comparison purposes, what third-party public sources (institutional popularization video, press release of 01/04/2026, preprint HAL hal-05560338) describe the Franco-Japanese “DNA cryptography” approach. Freemindtronic welcomes this research and reminds us that the technical objects differ from EviDNA (2024) and the cryptographic genome (2026).

What the corporate video exposes (non-empowering summary).

A Franco-Japanese team (Gulliver, CNRS/ESPCI Paris — PSL laboratory: Matthieu Labousse, Yannick Rondelez; XLIM, University of Limoges: Philippe Gaborit; partner University of Tokyo) presents cryptography by DNA as a new chapter in the The history of encryption.

1. Material. The DNA here is fully synthetic produced outside of any biological process. Four bases A, T, C, G form a “quaternary language” analogous to the binary (0/1): an ordered sequence encode information.
2. Cryptographic property sought. Synthesis is used to generate statistically random sequences — source of entropy for cryptography.
3. Encryption scheme. The protocol chosen is the (OTP — One-Time Pad): a random mask, as long as the message, used once; combined with the binary message to encrypt; recombined on the recipient side to decrypt. Theoretical safety is based on the randomness of the mask.
4. Role of the molecule (explicit video wording). The synthesized DNA molecule does not contain the message: it carries the future encryption key. Two identical samples are prepared (Tokyo / France demonstration); Each matching sequence their sample just before the communication to get the same binary key.
5. Operational chain. Sequencing (reading nanopore: differential current per base A/T/C/G) → software reading of the ATGC sequence → conversion to binary → encryption of the digital message in France → sending of the encrypted message (e.g. email) → decryption in Japan with the identical key.
6. Applications mentioned. Critical communications: defense, diplomacy, patents, financial exchanges; so-called “unconditional” security in the sense of OTP.

CNRS Operational Chain — Molecular OTP (text diagram, public sources).

 random synthetic DNA
        │
        ▼
Duplication ──► copy France ════ Japan copy
        │
        ▼  (just before the message)
Nanopore sequencing (×2) ──► IdenticalATGC sequence
        │
        ▼
ATGC → binary → OTP mask (|mask| = |message|)
        │
        ▼
Message ⊕ Mask ──► Channel (e.g. email) ──► Encryption ⊕ samemask

Advantages and disadvantages of Vernam encryption (literature review of a classical scheme, register A). The protocol adopted by the CNRS is based on the Vernam encryption (One-Time Pad), the properties of which have been established in the cryptographic literature since the work of Claude Shannon (1949). This reminder, which is unrelated to the Freemindtronic mechanisms, sheds light on the trade-offs of the institutional scheme.

Avantages.

• Perfect secret proved (perfect secrecy, Shannon): Under its three conditions, the cipher alone does not reveal none information about the clear message.
• Resistance to any computing power, including a future quantum computer: security is informational, non-computational.
• Simplicity of operation: The encryption is reduced to a bitwise XOR between message and mask.

Disadvantages (structural constraints).

• Key as long as the message: encrypting n bytes requires n bytes of mask — hence a storage and distribution cost proportional to the volume exchanged (the press release mentions messages up to several hundred megabytes, so as much key material).
• Strictly one-time use: Any reuse of a mask breaks the perfect secret (encryption correlation attack).
• Distribution and synchronization of the mask: both correspondents must have a identical and secret mask before the exchange — this is the central problem that the molecular chain (DNA duplication, physical transport, sequencing “moment T”) seeks precisely to solve.
• Perfect random required: Any statistical bias of the mask degrades the theoretical guarantee.
• Lack of intrinsic authentication and integrity: the Vernam cipher but does not prove the origin or non-alteration of the message; it must be supplemented by separate mechanisms (MAC, signatures).

These properties explain why the OTP, although theoretically optimal, remains operationally demanding and lends itself above all to punctual critical communications — a framework claimed by CNRS sources. They also shed light on the cross-reading of §1.6.1: a cryptographically monolithic scheme (an imposed mechanism) is opposed to an agnostic layer admitting several mechanisms depending on the policy.

Vernam Principle / OTP (text schema, classical cryptography).

Émetteur                              Destinataire
────────                              ────────────
clear message (M) encrypted message (C)
random mask (K) ── channel ──► samemask (K)
     │                                      │
     ▼                                      ▼
C = M ⊕ K                            M = C ⊕ K

Conditions: |K| ≥ |M|  ;  K used only once;  K perfectly random

Three “DNA” trajectories — distinct technical objects (text diagram).

         ┌──────────────────┬──────────────────────┬─────────────────────────┐
         │ CNRS 2026 │ EviDNA 2024 │ Genome / Digital DNA │
         │ (réf. externe)   │ (Freemindtronic)     │ 2026 (Freemindtronic)   │
├────────┼──────────────────┼──────────────────────┼─────────────────────────┤
 Source │ Synthetic DNA │ Human DNA Profile │ Procedural Generator │
 Secret │ Tube + Sequencing │ NFC + Paper QR │ TPM/vTPM + runtime │
 Crypto │ Vernam/OTP only │ mechanisms according to policy* │ PQC agnostic layer* │
 Time │ Instant T │ Enrollment + session │ T₀ → Tn (continuity) │
└────────┴──────────────────┴──────────────────────┴─────────────────────────┘
         * OTPs and other mechanisms according to policy — not imposed as a single scheme

What the CNRS press release (01/04/2026) adds. Preparation of duplicated DNA sets of synthetic origin; just before communication key generation by sequencing; Messages up to several hundred megabytes demonstration during the presidential trip to Japan; HAL title: Synchronized DNA sources for unconditionally secure cryptography (Jaudou, Gasnier, Boudjella, et al.).

Read-across (register A, without legal advice). The CNRS video confirms that the 2026 institutional approach is focused on molecular OTP: random synthetic DNA → Vernam mask → physical synchronization of two copies → point sequencing. EviDNA (2024) previously documented another invention: DataShielder Defense NFC HSM product using a human DNA profile (technical detail B/C register). The cryptographic genome and the ADN Digital (2024–2026) extend a third trajectory: time-trusted architecture, beyond the distribution of keys at a given time. The three axes share the word “DNA” but do not cover the same technical object. For the analysis of the generation of randomness and operational complexity respectively, see §1.6.1.

1.6.1. Random Generation and Operational Complexity — Comparative Reading (A-Register)

Purpose of this subsection. Check, using public sources only, whether the two trajectories use comparable of random generation and similar levels of operational complexity. This analysis does not constitute a value judgment on the scientific quality of CNRS work; It specifies distinct technical dimensions useful for cross-reading the dissertation.

What CNRS sources document (April 2026). The Franco-Japanese approach aims to solve a classic constraint of the OTP/Vernam: to produce and synchronize, between distant correspondents, a key perfectly random, as long as the message and single-use. To do this, researchers are mobilizing a molecular and instrumental chain:

1. Synthesis of entirely artificial DNA, whose order of bases A/T/C/G is statistically random;
2. Enzymatic duplication in strictly identical copies, kept at the sender’s and recipient’s premises;
3. Physical transport or pre-distribution of such samples;
4. Nanopore just before communication, on both sides, to read the same sequence;
5. Conversion ATGC → binary key → Vernam encryption of the digital message.

Two axes of complexity — non-interchangeable (text schema).

CNRS 2026                              Freemindtronic (ADN Digital / génome)
─────────                              ─────────────────────────────────────

OPERATIONAL COMPLEXITY OPERATIONAL COMPLEXITY
        ▲  ISLEVISE                              ▼  FAIBLE (post-config)
        │ lab · Sequencing │ Smartphone · TPM · runtime
        │ Physical transport │
        │                                      │
CRYPTO Complexity CRYPTO Complexity
        ▼ LOW (OTP only) ▲ HIGH(agnostic layer)
        │ Imposed Vernam │ Multiple mechanisms · continuity

Third-party sources (CNRS press release, IMT Atlantique, press popularization) also highlight biological and instrumental locks: sequencing noise, statistical bias in database pairing, the need to detect an interception of DNA material, sequencing machines and molecular biology protocols. At this stage, it is a proof of concept in a controlled environment, whose processing times are not intended for general public use on mobile devices.

What the Freemindtronic trajectory documents (Digital DNA/genome, registry A). The DNA Digital and the cryptographic genome do not use /strong<> molecular synthesis or biological sequencing. The expression “DNA” here refers to a procedural metaphor: an organization of trust inspired by the structural principles of the living genome (segmentation, inheritance, continuity, reevaluation over time) — without exploitation of biological DNA or DNA computing (see EviSKMS memory §29.6 on the authentication of living beings).

In this trajectory, the generation of random or pseudo-random material for the trusted identity is done by a procedural generator integrated with the cryptographic genome and governed by the EviSKMS/CryptPeer runtime. The internal mechanisms of derivation, genomic transition and digital DNA correlation → segments fall under the C register; in the A register, only the operating result is documented: after the initial import of the certificates, the usage becomes transparent for the operator (§1.7).

Comparative synthesis — two axes of complexity, not interchangeable.

Documentary conclusion (register A). The CNRS approach is operationally more demanding (molecular infrastructure) and cryptographically monolithic: the public protocol retains only Vernam/OTP. Freemindtronic’s DNA Digital / genome trajectory is based on a software architecture that can be industrialized, capable of producing OTP</strong keys> when the policy requires it, without limitation — and mobilizing other cryptographic bricks according to the governance policy, in a logic of continuous trust beyond the mere distribution of masks at a given time. For a mapping of the other global “DNA + security” families, see §1.6.2.

1.6.2. International mapping — “DNA + security” families and Freemindtronic distinction (Registry A)

Status. This subsection does not claim authorship on the third-party works cited. It synthesizes, from public sources (journals, preprints, research programs), a documentary taxonomy useful for locating the Freemindtronic trajectory (EviDNA, ADN Digital, cryptographic genome, CryptPeer/EviSKMS) in the face of all the global research mobilizing the “DNA” and “security” couple — including cyber, storage and molecular cryptography.

Observation Two recent syntheses (IEEE Access, 2023; iComputing, 2024) converge: the field is fragmented, poorly standardized, and often mixes — in the literature — real molecular approaches, software simulations inspired by DNA, and structural metaphors. The word “DNA” thus covers several non-interchangeable technical objects — which this thesis formalizes to avoid any confusion of authorship or reproducibility.

Seven documentary families (text schema, register A).

F1 Molecular OTP / Synchronized Entropy CNRS 2026 · ANR DNA Sec (in progress)
F2 Origami / Structural Nano Cryptography Zhang 2019 · 3D extensions (lab)
F3 Molecular Steganography Clelland 1999 · NAPDISS 2024 (Cover-Up)
F4 Pseudo-DNA software many articles · especially simulation
F5 DNA Storage + Hybrid Encryption Noise Channels · Massive archiving
F6 DNA Database Security DNA Sec Program (Theft · Tampering)
F7 Freemindtronic Procedural Genomic Cryptography 2018–2026 (≠ molecule)

Read-across matrix — dimensions that distinguish F7 (Freemindtronic).

Indirect valuation (Ledger A, no legal opinion).

• Functional coverage. The F1–F3 families cover perfect secret distribution, structural nano and concealment, respectively. None of them publicly documents, to date, an industrialized continuous trust architecture on a terminal — the object of F7.
• OTP without exclusivity. F1 demonstrates the institutional interest of molecular OTP; F7 can use the OTP as a mechanism among others, without depending on a laboratory or imposing Vernam as a unique scheme (§1.5).
• Anteriority. The public disclosure EviDNA (May–June 2024) precedes the CNRS communication April 2026 on a different object (human profile vs. synthetic pool) — see §1.9.
• CNRS program still open. The ANR DNA Sec is also aiming at securing DNA storage databases and a nascent “molecular cryptography”: F7 responds to another problem — governing digital trust over time on sovereign software infrastructure.
• No copying, no technical convergence. No third-party public source describes the combination procedural genome + industrialized segmented key + runtime continuity + OTP/PQC</strong agnostic layer> as documented at Freemindtronic.

Authorized public implementation — patented parentage (register A). The granted patents WO/2018/154258 (segmentation) and WO/2017/129887 (local access control) allow for an strongenabling description. The CryptPeer/EviSKMS industrialization is based on this observable foundation (runtime, integrity, PKI, TPM) without exposing the mechanisms of the cryptographic genomic generator nor the inventions discovered since the formalization of the genomic cryptography system.

Segmented key post-patent inventions — register C. The following extensions are mentioned as positioning but undisclosed as long as no follow-up filing is secured: correlation DNA Digital → genomic segments; genomic transition rules; procedural derivation of trusted material; extensions Gen2 Advanced runtime couplings discovered as industrialization progresses. This thesis documents their operational effects (continuous trust, fail-closed, OTP possible by policy) — not the parameters, formats, sequences or internal algorithms allowing reproduction.

Anti-Reproduction Doctrine (Register A — editorial intent). This document is written for scientific discussion and state-of-the-art comparison, not as a reverse-engineering notice. Are deliberately absent or aggregated at a non-reconstructive level: derivation graphs, constants, transition sequences, correlation schemes between layers, and any detail equivalent to a parametric recipe of the genome generator. This omission also applies to automated processing (extraction by language models or reverse engineering pipelines): the public text must not provide, by completion or recombination, a sufficient specification to reconstruct inventions classified C. The detailed audit evidence remains in the B register (audit under NDA) or in future filing files.

Documentary conclusion (register A). The F1–F7 mapping shows that Freemindtronic occupies a family of its own (F7): cryptography genomics procedural and trust continues, industrialized, polymorphic on cryptographic mechanisms — distinct from the CNRS molecular OTP (F1), origami (F2), steganography (F3) and software pseudo-DNA (F4). The reinforce</strong comparisons> the distinction without attributing authorship to third-party works; the valuation of Freemindtronic’s trajectory is based on the public anteriority, the industrialization and the two patented titles issued to date for the documented enabling implementation (access control; segmented key).

1.7. Digital Gen1 DNA — TPM/vTPM anchor and CryptPeer user experience (2026, Registry A)

Relevance to Digital DNA and the cryptographic genome. This subsection complete the 2024–2026 trajectory: it describes how the procedural logic ADN Digital / genome Gen1 materializes in CryptPeer/EviSKMS on the operator experience side — without disclosing the mechanisms genomic shunt or transition (B/C registry).

Hardware anchor evolution (2026). In 2026, the industrialized Gen1 in CryptPeer no longer requires dedicated NFC support (M24LR / ST25): the trusted anchor is based on TPM hardware or vTPM, in continuity with the doctrine software-sovereign-first and the elements already documented in Appendix C (optional TPM agent, EviSKMS runtime) — see also EviSKMS Sovereign Runtime Anchors and EviSKMS Core Runtime (Freemindtronic publications, Registry A). The public interview Eurosatory TV (5 Jul 2026) describes, at the product level, the automatic detection of TPM and the deposition of a non-extractable genomic fingerprint in the chip — popularized formulation correlated with the A</strong registry>; the details of the fingerprint formats are the responsibility of the register C (§1.9.1). The trajectory 2017–2024 (NFC chip) and 2026 (TPM/vTPM) illustrates a generalization: from point-in-time hardware evidence to a time-governed runtime trust.

CryptPeer User Experience (Registry A, Product Level).

After the initial import of the certificates on the terminal, CryptPeer allows transparent use in 100% local mode; in remote mode, transport relies on Let’s Encrypt in a server blind model where the content remains end-to-end encrypted.

CryptPeer Modes of Exploitation (Text Schema, A Register).

                    ┌── Import initial certificats (friction unique)
                    ▼
              Approved Terminal
                    │
        ┌───────────┴───────────┐
        ▼                       ▼
  100 % sovereign-local    Mode distant
  E2EE · passwordless      TLS Let's Encrypt
  Transparent Blind Server Runtime (E2EE)
        │                       │
        └───────────┬───────────┘
                    ▼
        Confiance continue Gen1 (TPM/vTPM · DDNA · RI)

Limits (Registry A). Correlation details DNA Digital → genomic segments → TPM/vTPM anchor, internal formats, and transition rules fall under the C registry. This paragraph does not constitute a reproduction notice. For the published infrastructure layer (doctrine, PKI, anchors, runtime integrity), see §1.8.

1.8. EviSKMS Technology Publications (Freemindtronic.com, Register A)

Freemindtronic has published on its website four technology pages which complete this thesis on the trajectory DNA Digital / Gen1 genome / CryptPeer — without replacing the evidence appendix or disclosing any enabling mechanism (C registry). They articulate the sovereign doctrine, the PKI evidence-bound, the anchor runtime (TPM) and the integrity runtime — pillars of industrialization 2026.

Read-across memory ↔ site. The dissertation formalizes the scientific framework and the trajectory DNA / genome; Freemindtronic pages detail the industrialized sovereign trust infrastructure. Together, they document the continuity DataShielder (NFC, 2017–2024) → CryptPeer/EviSKMS (TPM, genome, 2024–2026).

1.9. Public Sources of Disclosure and Anticipation

This section lists time-stamped public disclosures prior art of Freemindtronic inventions — cryptographic genome, ADN Digital, EviDNA, segmented trust — without duplication of enabling mechanisms (A registry only). The common thread is the inventive trajectory (2018 patent → CryptPeer implementations → industrialization); The videos and web publications below are the correlated public proofs. Defense fairs (Eurosatory, etc.) are cited as contexts of disclosure, not as the main subject of the dissertation.

1.9.1. Interview Eurosatory TV — cryptographic genome (5 July 2026, register A)

Source and rights. Public interview broadcast on the YouTube channel Eurosatory: https://www.youtube.com/watch?v=amwVAGp9LHw — Jacques Gascuel (Freemindtronic SL) and David Amsellem (AMG PRO, distribution). English subtitles (SBV lounge). This synthesis cite and structure public statements; it does not constitute not an enabling record beyond the A register. It sets out the documentary correlation between the oral disclosure at the fair and the present thesis (copyright on the inventor’s formulation; work of formalization protected).

Objet. Verify, after public broadcast, that the interview remains aligned with the formalized trajectory of the dissertation — segmentation, trust over time, ADN Digital, CryptPeer — and specify what is not disclosed (internal mapping, generator parameters, detailed DDNA formats: registry C).

Chronological synthesis (public statements).

Technical topics — read-across register A.

Formulations to be nuanced. “Impossible to falsify”, “inviolable” or “end of cyberattacks” are part of the vulgarisation salon. The brief translates them into falsifiable terms: segmented trust, fail-closed, attack surface reduction — with no absolute guarantee. See Limits and falsifiability.

Out of scope (register C). Internal mapping, generator algorithms, detailed DDNA formats, ASC modules — §1.12.

Documentary conclusion. The interview publicly confirms the 2024 pivot → 2026 and the focus on segmentation and confidence over time — without reproduction instructions. Bibliography: Eurosatory TV 2026.

1.10. EviDNA Proof of Implementation — DataShielder Defense NFC HSM (Registry A)

The commercial base (encrypted QR + NFC, without DNA) is marketed since 2017 on M24LR 64K NFC (STMicroelectronics). Between 2022 and 2024, Freemindtronic is adding the ST25 64K NFC compatibility and the layer EviDNA (human DNA profile → keys). The Defense with human DNA is publicly disclosed in 2024 (web, videos — §1.9). Between 2024 and 2026, the trajectory extends into ADN Digital and cryptographic genome (CryptPeer/EviSKMS).

Material filiation (register A).

Public Proof of Anteriority (Registry A). The demonstrations and publications of May–June 2024 (§1.9) establish the existence of a product DataShielder Defense NFC HSM mobilizing a human DNA profile for cryptographic trust, without this brief reproducing the detailed technical chain (derivation, encapsulation, sharing) — this is the responsibility of the B/C registry as long as no additional repositories are secured.

What Registry A allows to formulate. Commercial product; NFC hardware support (M24LR / ST25); EviDNA layer publicly documented in 2024; accesscontrol architecture to protected memories (WO/2017/129887) and segmented key (WO/2018/154258); field use without molecular infrastructure. What remains unpublished: derivation parameters profile → trusted material, internal formats, detailed sharing schemes, encrypted QR capabilities, code module names.

Source anchor — two evidentiary registers.

Important (Registry A). A GitHub repository private is not a public disclosure in the patent sense: it does not replace public sources (web, video, memory), but reinforce the proof of implementation in the B registry.

The detailed implementation (code structure, modules) falls under the B register. Explicit limits (register A). The public anteriority is based on the demonstrations and publications of 2024, prior to the institutional announcements of 2026; the detailed proof of implementation (private repository, commits, code) falls under the B registry.

Distinction vs CNRS 2026 (registry A). EviDNA mobilizes an imported human DNA <> as a trusted material for encryption and signature (B/C registry detail) — it is not nor a pool of duplicated synthetic DNA, ni a molecular OTP synchronization “just before the message” as described by the CNRS. The cryptographic genome (2026) extends this trajectory towards a trust governed over time; it can produce OTP</strong keys> depending on the governance policy, without limitation to this scheme — beyond the point-in-time identity “it’s me” at time T (§1.5).

Distinction méthodologique 2024 / CNRS 2026 / Freemindtronic 2026. The milestone EviDNA (2024) documents a implemented invention: DataShielder Defense NFC HSM product (technical detail registry B/C), with public disclosure by time-stamped videos (§1.9). The CNRS communication of April 2026 describes a distinct approach (synthetic DNA, OTP/Vernam, HAL hal-05560338). The 2026 Freemindtronic milestone documents the Digital DNA and the cryptographic genome in CryptPeer/EviSKMS. Gen2 is implemented in CryptPeer; mechanisms detailed in register C.

Perceived proximity and risk of confusion. Reading institutional press releases, listening to interviews or watching videos, the public can perceive a strong semantic proximity between “DNA” and “cryptography”. This media proximity must not lead to confusion of authorship or to the absorption of previous inventive trajectories — in particular the cryptographic genome, which aims at a trust continuous over time, distinct from the identity punctual at the time T (“it’s me” at the time of authentication or the generation of OTP keys). See §1.5. For the canonical definition of EviDNA, its direct comparisons and its patented parentage, see §1.11.

1.11. EviDNA — technical object, patented parentage and direct comparisons (A</h4 register> Purpose of this section. Centralize, at a non-enabling level, everything that specifically concerns the invention EviDNA (2024): definition, stacking with the segmented key patent, operator pathway, comparisons with the neighboring state of the art, bridge to Digital DNA (2026), limits and regulatory positioning. The internal mechanisms of derivation profile → trusted material fall under the B/C register.

1.11.1. Canonical definition — what EviDNA is (and what it is not)

EviDNA refers to the Freemindtronic layer (public milestone May–June 2024) that mobilizes an imported human DNA profile — a structured file provided by the operator — as trusted material to produce cryptographic material (encryption, signature; mechanisms according to policy — detail registry B/C). It is industrialized in the product DataShielder Defense NFC HSM, on an encrypted QR pad + NFC</strong token> (STMicroelectronics M24LR / ST25).

Sub-milestone in the F7.</strong family> In the mapping §1.6.2, EviDNA is the sub-milestone “human profile + NFC product”; DNA Digital / genome (2026) is the procedural generalization without breaking philosophy (materialized trust, not molecule).

1.11.2. Patented parentage and technical stacking (register A)

Patented stack — three separate layers (A register).

The industrialization DataShielder (M24LR / ST25, including Defense) combines layer access control (conditional opening of the chip’s protected memories via NFC token terminal ↔ local link) and layer segmentation (154258). Other wireless protocols local (Wi-Fi, Bluetooth, etc.) can extend the sameprinciple depending ondeployment; the NFC mode is the documenté for EviDNA 2024 (§1.10).

2016-2020 WO/2017/129887 — Access control · Local wireless · Protected memory
2018-2019 WO/2018/154258 — segmented key · Proximity · NFC token
        │
2017 ─────┴──► Encrypted QR Base + M24LR NFC (Commercial, DNA-Free)
        │
2022-24 ───► ST25 Compatibility +EviDNA Layer Development 
        │
2024 ──────► EviDNA: Human DNA Profile → Trusted Material
        │         DataShielder Defense NFC HSM
        │
2024-26 ───► Digital DNA + giscryptographique name (gisnisralisation)
        │
2026 ──────► CryptPeer/EviSKMS · TPM/vTPM (NFC non obligatoire)

The EviDNA layer does not replace patents: it stacks on the access control + segmentation base. No parametric correlation profile → segments is published here.

1.11.3. Operator journey — “three gestures” (register A)

Publicly documented (videos §1.9, press sheet): smartphone + paper + NFC chip. The secret of the reconstruction does not lie on paper: the encrypted QR allows remote sharing (email, display) while the hardware key remains on the NFC token only (physical proximity — patented principle).

 Legitimate Operator
     │
     ├─► QR scan (paper or screen) ──► no raw secrets on paper
     │
     ├─► NFC approach (M24LR / ST25) ──► conditional reconstruction (patent)
     │
 └─► Costed/ signed  session ──► memechanisms according to policy (B/C)

Paper printing (A register). A4 support with multiple encrypted QRs; The 2024 press release and demonstrations document a without exposing the secret on paper exchange capability — consistent with the segmented patent doctrine.

1.11.4. Comparison — encryption/computation on genomic data (Registry A)

Another branch of research protects the genomic file itself (cloud storage, homomorphic computation, allele masking) — EviDNA’s distinct object, which uses a profile as crypto trust material, not as a hosted medical database.

1.11.5. Comparison — live biometrics and point identity (register A)

Freemindtronic does not use FIDO as a foundation of trust (§1.4); The table above is an external literature comparison, not an interoperability claim.

1.11.6. Pont EviDNA (2024) → ADN Digital / genome (2026)

EviDNA is not obsolete: it remains the documented founding milestone (prior 2024, video evidence) of the F7 lineage; ADN Digital is the industrialized generalization (§1.7).

1.11.7. Regulatory context, use cases and EviSKMS link (Registry A)

Genetic data (without legal advice). The GDPR treats genetic data as special category (art. 9). EviDNA does not claim not the massive hosting of genomes in the cloud: the profile is mobilized under operator control on terminal and token, in line with a sovereign local logic — distinct from DTC models (consumer tests) whose leaks have illustrated the risks of centralization.

Publicly Documented Use Cases.

• Defense / counter-espionage — public primer 2022 (defense exhibition); version Defense Eurosatory Lab May 2024 (Freemindtronic announcement).
• Sensitive exchanges — encryption and authentication with portable trusted hardware (NFC + QR).
• Remote sharing — Encrypted QR without carrying a molecule or key in plain text on paper.

EviSKMS Memory Link. The authentication of living beings — presence, life, context (EviSKMS memory §29.6) deals with the living/artifact distinction; EviDNA, on the other hand, treats the imported profile as a trusted material produced — complementary axes, objects not confused.

1.11.8. Specific limits EviDNA (Registry A)

• EviDNA does not provide molecular OTP or perfect informational secrecy in the Shannon sense of the CNRS.</li protocol>
• It does not constitute a genomics research platform, GWAS cloud or homomorphic computation on third-party genomes.
• It does not replace medical advice, genetic diagnosis or civil identity eIDAS.
• The quality and provenance of the imported profile are the responsibility of operator governance (outside the public technical perimeter).
• The dedicated falsifiable hypotheses are in § Limits — EviDNA component; the derivation mechanisms remain in the register C.

Synthesis (Registry A). EviDNA is the Freemindtronic invention that set the first public milestone of cryptography mobilizing a human DNA profile as a trusted material on commercial product, before the molecular OTP (2026) and distinct of encryption of genomic files. Its documented public implementation is based on the key</strong patent>; Its genomic extensions are part of future deposits. For the framework of assumed non-disclosure (including CryptPeer), see §1.12; For competitive reading and renowned laboratories, §1.13.

1.12. Controlled publication — upcoming complementary patents and CryptPeer scope (register A)

Status. This section explains, in scientific language, why the does not disclose everything — including the implementation in CryptPeer/EviSKMS. This is not an unintentional omission, but a methodological choice related to the protection of intellectual property in the process of being secured.

Principe. As long as complementary inventions (EviDNA detailed, Digital DNA, genome generator, Gen2 extensions, advanced runtime couplings) are not securely deposited, any enabling publication would risk anticipating the state of the art and weakening residual IP. The dissertation thus adopts a posture of non-reproducible scientific discussion: it establishes the problem, the trajectory, the distinctions, the proofs of maturity and the limits — without providing the parameters allowing a reconstruction.

CryptPeer Perimeter (Registry A). The industrialization CryptPeer/EviSKMS is documented as proof existence and maturity runtime: integrity, evidence-bound PKI, TPM anchors, sovereign passwordless, DRT continuity, test campaign — without genomic core reproduction instructions. The reader can verify that a product exists and works; he cannot, from the dissertation alone, reconstruct inventions classified C. This frontier also applies to automated processing (LLM, assisted reverse engineering).

Closing Wording (Register A). As it stands, the granted international patents WO/2018/154258 and WO/2017/129887 allow for a public description enabling at the architectural level (segmentation; local access control). The derivation EviDNA and the genome remain attested (product, videos, industrialization) but not fully published — pending IP security. This reservation will be gradually lifted by controlled deposits and complementary publications (§1.2).

1.13. Competitive landscape, renowned laboratories and indirect valorization of EviDNA (Registry A)

Objet. Situate EviDNA in relation to the solutions and laboratories which, by their reputation and advancement, structure the “security + DNA / genome” market — without any claim of absolute superiority or legal opinion. The desired effect is a enhancement by documentary contrast: the more credible and active the adjacent state of the art, the more readable the distinct technical object of EviDNA becomes.

Constat. No identified public source documents, to date, the following combination: human DNA profile imported → operational trusted material→ segmented key HSM NFC token → QR encrypted without secrets on paper → commercial product disclosed in 2024. Renowned players mainly deal with other problems — protection of genomic files, OTP molecular, or centralization DTC — which, through intellectual capitalarity, strengthens EviDNA’s positioning rather than weakening it.

Indirect valuation reading (register A).

• Scientific capital effect. The activity of prestigious laboratories (CNRS/ESPCI, CISPA, Columbia/Broad, EPFL, Genome Research) confirms that the “genome + security” boundary is strategic — but according to technical objects different from that of EviDNA.
• No documented direct competition. None of the players mentioned publicly claims the same product stack (human profile + segmented NFC key + QR + 2024 defense field use).
• Apparent complementarity. Cloud/HE searches could coexist with a operational trust layer on the terminal — objects not merged in this thesis.
• Enhanced Anteriority. The EviDNA disclosure May–June 2024 precedes several recent public milestones (CNRS 2026, SQUiD 2024 in archiving) on related but not identical problems.

Limitations of this analysis (Register A). The table is not intended to be a comprehensive systematic review; It selects representative and verifiable references to inform positioning. The absence of an actor in the table does not mean the absence of related works not cited. Freemindtronic does not minimize the quality of third-party searches; it specifies the non-recouvreance with the EviDNA object.

Synthesis (Registry A). The global landscape validates the importance of the subject while showing that EviDNA occupies a niche of its own: trusted material derived from a human profile, industrialized, anchored on a segmented key patent — beyond genomic storage, homomorphic cloud and molecular OTP. This reading completes the thesis for a documentary closure of the comparative component. For the “genomic privacy” research ecosystem (iDASH, Beacon), see §1.14.

1.14. Genomic privacy — iDASH, Beacon (Broad/Stanford) and scientific equity (Registry A)

Objet. Complete §1.13 by the research on the sharing and re-identification of genomic data — a field that has been structured for more than fifteen years (MIT, Stanford, Broad Institute, Columbia, NIH/iDASH).

Historical observation. As early as 2008, Homer et al. showed that it was possible to infer the presence of an individual in an aggregated dataset (bib.). The Beacon (GA4GH) network enabled binary queries on research cohorts. In 2015, Shringarpure and Bustamante (Stanford) demonstrated re-identification attacks on these services (bib.). The iDASH Genomic Privacy & Security Workshop 2016 devoted tracks to Beacon mitigation and computation on encrypted genomes (bib.).

Capitalarity (Registry A). The intensity of genomic privacy research confirms the strategic importance of genetic data (GDPR art. 9, §1.11.7). No work cited documents the stacking produced EviDNA (2024). iDASH and Beacon indirectly reinforce its valuation by showing the limits of centralized or federated sharing models.

1.15. Roadmap for future publications (Register A)

Status. What can be published after securing PI — without a timetable commitment. Complete§1.12.

Principe. Each phase expands the public register without transforming the memoir into a reproduction record. CryptPeer remains attested in phases 2–3 as proof of maturity runtime.

EviDNA cryptography — Limits, falsifiability and scope of validity

What this memoir doesn’t pretend to prove

• An independent security audit or a certificate of compliance (eIDAS, Common Criteria, FIPS);
• A published quantitative benchmark opposing EviSKMS to FIDO or PKI in all contexts;
• An enabling technical notice allowing the reproduction of Gen2 or detailed EviDNA mechanisms (C registry);
• An equivalence between the Freemindtronic procedural randomness and the CNRS molecular OTP perfect randomness;
• Clinical or regulatory validation of the use of imported DNA profiles (EviDNA) beyond documented product demonstrations;
• A substitution for a cloud genomics vault (PROMISE, Varlock, etc.) — separate search object (§1.11.4).

Falsifiable hypotheses — EviDNA (2024)</h3 component> H-E1 — NFC Segmentation and Proximity. Utterance. Without an approved NFC token and physical proximity in accordance with the patented model, trust reconstitution for an EviDNA session fails (denied or no operation). Rebuttal. Successful session with QR only, with no expected token present.

H-E2 — Absence of paper secrets. Statement. Inspection of the paper medium (printed QR) does not allow the reconstruction of the trusted material equivalent to the NFC token. Refutation. Extraction of complete secrecy from paper alone, reproducible on documented sample.

H-E3 — Uniqueness of the trusted material. Statement. Two distinct DNA profiles, under the same product policy, do not produce an interchangeable trust material (black-box test on observable outputs). Refutation. Collision or interchangeability demonstrated without knowledge of the internal mechanism.

H-E4 — Distinction vs. Molecular OTP. Statement. EviDNA does not require nanopore sequencing or molecular sample duplication for a documented session. Réfutation. Molecular instrumental dependence identical to the CNRS protocol on the same product scope.

H-E5 — Anteriority product. Statement. The time-stamped public sources of May–June 2024 precede the CNRS communication April 2026 on a separate technical object. Rebuttal. Third-party public source establishing a prior disclosure of the same object (human profile + NFC HSM + QR) by another actor.

Falsifiable hypotheses — digital trust component (EviSKMS Gen1)

H-C1 — Continuity vs. point-in-time authentication. Utterance. A segmented trust architecture that is re-evaluated over time, and governed at runtime, reduces spoofing scenarios compared to point-in-time, comparable friction MFA. Refutation. Lack of measurable gain on a predefined battery of scenarios.

H-C2 — Fail-closed runtime. Utterance. If runtime integrity or continuity regression is detected at startup, the system denies exploitation. Rebuttal. Exploitable without alert after controlled corruption of continuity artifacts.

H-C3 — DDNA Gen1 without raw data exposure. Statement. The Gen1 foundation allows traceability by standardized fingerprints without transit of sensitive raw sequences. Refutation. Reproducible leakage of raw data in transit or logs.

H-C4 — Multi-surface anti-replay. Utterance. Anti-replay guardrails prevent successful reuse of queries that have already been consumed. Refutation. Successful replay attack on a qualified surface.

H-C5 — Documented differentiation vs. standards. Statement. EviSKMS Gen1 provides measurable value on at least two criteria of the comparative table §1.4. Refutation. No favorable deviation observable on the tested perimeter.

EviDNA DNA cryptography: PI</h3 constraint> The publishing strategy (A/B/C registries) strengthens IP protection but reduces immediate external tamperability on mechanisms classified C. See §1.2 and the mapping §1.6.2.

Publicly cited issued titles. The patents WO/2018/154258 (segmented key) and WO/2017/129887 (access control) constitute the two granted titles on which the dissertation can rely for an enabling architecture description. All inventions related to genomic cryptographic generator, detailed EviDNA, ADN Digital, extensions Gen2 and discoveries subsequent to the creation of the genomic cryptography system are included in the C register until further deposit.

Publication vs reverse engineering. The dissertation values observable results (product, runtime, comparisons, anteriority) and public patented filiation, without providing a reconstructive specification of the genomic core. This rule also applies to automated uses (LLM, code extraction, assisted reverse engineering): the A register text must not be sufficient, alone or recombined, to deduce internal parameters, transitions or derivations. Detailed evidence is reserved for the B (NDA) registry or intellectual property files in preparation.

CryptPeer and upcoming patents. Implementation in CryptPeer/EviSKMS is attested at the non-enabling level: architecture, functional effects, evidence of industrialization — not the internal mechanisms of segmented key post-patent inventions. This boundary is explained in §1.12. It does not indicate a deficiency in the memory, but a waiting for PI to be secured before any further disclosure.

Conclusion

This thesis establishes that the Freemindtronic trajectory (EviDNA 2024, ADN Digital, cryptographic genome 2026, CryptPeer/EviSKMS) constitutes a distinct object from recent institutional approaches on synthetic DNA and OTP/Vernam (CNRS 2026), while saluting the corresponding academic research.

It documents an industrialization observable (Gen1/Gen2 in CryptPeer) at a non-enabling level, a patented parentage (WO/2018/154258), the canonical definition EviDNA (§1.11), a controlled publication doctrine (§1.12), a international map, a competitive landscape (§1.13), the ecosystem genomic privacy iDASH/ Beacon (§1.14) and a roadmap complementary publications (§1.15).

GDPR positioning (register A, without legal advice). genetic data falls under the Article 9 of the GDPR (special category). EviDNA is part of a logic of minimization and local control by the operator: profile imported as a trusted material on an approved terminal/hardware, without cloud centralization comparable to DTC players (§1.13). Purpose, security (Art. 5 and 32) and impact assessment (Art. 35) remain the responsibility of the data controller — see §1.11.7.

The broader framework — predictive AI, agentic memory, cyber-physical trust — is developed in the EviSKMS reference memory.

EviDNA DNA cryptography — Selected bibliography

Entries cited in this memoir. Full IA bibliography: EviSKMS memory.

Gascuel, J. — Système de contrôle d’accès / Access Control System (2016–2020).

Links: WO/2017/129887 · FR3047099 B1 · EP3408777 Usage: standalone memory/protected device access control; local wireless communication (documented NFC); DataShielder NFC HSM stacking — §1.11.2 · §1.10.

Gascuel, J. — Segmented Key Authentication System (2018–2019).

Links: WO/2018/154258 · FR3063365 B1 Usage: patented parentage, segmented key, conditional trust reconstruction, variant jamming module (§1.1.1).

NIST SP 800-63-4 — Digital Identity Guidelines.

Links: NIST Usage: identity and authentication framework, external comparison.

NIST SP 800-207 — Zero Trust Architecture.

Liens : NIST Usage : comparaison cadre Zero Trust.

FIDO Alliance — Passkeys.

Links: fidoalliance.org/passkeys Usage: WebAuthn/FIDO external comparison (Freemindtronic does not use FIDO as a base).

W3C — Web Authentication Level 3.

Liens : W3C WebAuthn Usage : comparaison externe authentification forte.

ETSI EN 303 645 — Cyber Security for Consumer IoT.

Usage: comparison of IoT and connected objects.

EU Cyber Resilience Act (2024).

Usage: regulatory framework for connected products.

OWASP Top 10 for LLM Applications (2025).

Usage: AI threat context and continuous trust.

Eurosatory TV (2026) — Interview Jacques Gascuel, cryptographic genome and CryptPeer.

Links: YouTube amwVAGp9LHw Usage: public disclosure salon (5 Jul 2026); segmentation; confidence in time; Digital DNA; TPM; synthesis register A §1.9.1 — without enabling reproduction.

CNRS / HAL hal-05560338 (2026) — Synchronized DNA sources for unconditionally secure cryptography.

Links: HAL hal-05560338 Usage: CNRS external reference — OTP/Vernam, synthetic DNA; documentary comparison without claim of authorship.

Survey — DNA-Based Cryptography and Steganography (IEEE Access, 2023).

Links: doi.org/10.1109/access.2023.3324875 Usage: natural taxonomy / pseudo-DNA / steganography; Framework§1.6.2.

A Review of DNA Cryptography (iComputing / Science Partner J., 2024).

Links: doi.org/10.34133/icomputing.0106 Usage: state of the art, lack of standardized protocols; distinction F4 vs F7.

Zhang et al. — DNA origami cryptography for secure communication (Nature Communications, 2019).

Links: doi.org/10.1038/s41467-019-13517-3 Usage: F2 family — structural nanocryptography; indirect comparison.

ANR — DNA Sec : DNA data and Cybersecurity (ANR-24-CE39-3908).

Links: anr.fr · IMT Atlantique DNASec Usage: current F1/F6 program; Context Franco-Japanese research.

PROMISE — Controlling my genome with my smartphone (2021).

Links: doi.org/10.1007/s00392-021-01942-8 Usage: comparison of cloud genomic encryption + smartphone; distinction vs EviDNA (§1.11.4).

Varlock — Privacy-preserving storage of sequenced genomic data (BMC Genomics, 2021).

Links: doi.org/10.1186/s12864-021-07996-2 Usage: masking and confidential storage of sequenced genomes; separate object of EviDNA.

GDPR — Regulation (EU) 2016/679, Art. 9 (genetic data).

Links: EUR-Lex 32016R0679 Usage: special category frame; cautious positioning EviDNA (§1.11.7) — without legal advice.

Blindenbach et al. — SQUiD: ultra-secure storage and analysis of genetic data (Genome Biology, 2024).

Links: doi.org/10.1186/s13059-024-03447-9 Usage: HE / genomics cloud; distinction vs EviDNA (§1.13).

Huang et al. — GenoGuard: Protecting Genomic Data against Brute-Force Attacks (IEEE S& P, 2015).

Liens : doi.org/10.1109/sp.2015.34 Usage : honey encryption biobanque ; objet distinct stockage long terme.

TX-Phase — Secure phasing of private genomes in a trusted execution environment (Genome Research, 2025).

Links: genome.cshlp.org/content/35/12/2626 Usage: TEE and genomic pipeline; indirect comparison §1.13.

Homer et al. — Resolving individuals contributing trace amounts of DNA (PLoS Genetics, 2008).

Links: doi.org/10.1371/journal.pgen.1000167 Usage: genomic re-identification; §1.14.

Shringarpure & Bustamante — Privacy leaks from genomic data sharing beacons (AJHG, 2015).

Links: doi.org/10.1016/j.ajhg.2015.09.010 Usage: Beacon attack; §1.14.

iDASH — Genomic Privacy & Security Workshop 2016.

Links: humangenomeprivacy.org/2016 Usage: genomic privacy benchmarks; §1.14.

GA4GH — Beacon API.

Links: docs.ga4gh.org/beacon Usage: genomic federated sharing; separate from EviDNA (§1.14).

Glossaire

This glossary sets out the vocabulary of this thesis (EviDNA, Digital DNA, cryptographic genome) without constituting a reproducing-enabling record.

Appendix A — Synthetic prior art chronology (register A)

Objet. Legal reading and press at a glance — synthesis of §1.9 without enabling reproduction.

Lecture. Trajectoire salon : Eurosatory 2022 (projet) → 2024 (Defense industrialisée) → 2026 (CryptPeer). Filiation continue 2017 → 2026.

related documents

• EviSKMS memory — predictive intelligence architectures
• Press Release — Digital DNA and Cryptographic Genome
• White Paper