Skip to content

James Scott Institute for Frontier Sciences

Cybersecurity & Cyber-Physical Systems Research

Adversarial intelligence for the security, resilience, and scientific integrity of complex systems.

The James Scott Institute for Frontier Sciences extends its evidence-governed research methods into cybersecurity, cyber-physical systems, quantum security, AI assurance, and cyber conflict studies. Through KRYOS Hypercube Intelligence and its associated framework ecosystem, qualified research teams can investigate system vulnerabilities, model defensive resilience, challenge technical assumptions, and develop independently reviewable research findings.

Apply for a Research-Access AwardExplore the 24 Research Services

Selected research-access awards are personally funded by James Scott and administered by the Institute for Critical Infrastructure Cybersecurity. No financial assistance is provided.

Cybersecurity is one interdisciplinary research area within the Institute's broader frontier-sciences research programs.

Scientific connection

Where Frontier Science Meets Cybersecurity

Cybersecurity and physics are different disciplines. They meet where measurement, control, computation and adversarial pressure act on the same systems. KRYOS provides a common evidence, modeling, uncertainty, simulation, and adversarial-assurance architecture across both.

  • 01

    Scientific instrumentation and data integrity

    Measurements are only as trustworthy as the firmware, calibration chains and data paths that carry them.

  • 02

    Cyber-physical systems and control theory

    Control loops, interlocks and physical constraints decide whether a digital fault becomes a physical consequence.

  • 03

    Quantum computing and cryptographic security

    Quantum information science changes cryptographic assumptions and requires careful, evidence-bounded transition planning.

  • 04

    Artificial intelligence and adversarial robustness

    Models and agents face manipulation, poisoning and tool misuse that need falsifiable evaluation, not assurances.

  • 05

    High-performance computing and research infrastructure

    Scientific software, HPC environments and laboratory networks carry dependencies that shape research validity.

  • 06

    Complex systems, strategic uncertainty and cyber conflict

    Interdependent infrastructure and adversarial decisions call for uncertainty-aware models and preserved dissent.

Governed research flow
  1. 1Authorized Research Question
  2. 2Evidence and Provenance
  3. 3KRYOS Framework Selection
  4. 4Scientific/Cybersecurity Modeling
  5. 5Adversarial Simulation and Challenge
  6. 6Independent Validation
  7. 7Governed Research Findings

Six programs

Six Specialized Cybersecurity Research Programs

01

Cyber-Physical Systems and Critical Infrastructure

Protecting measurement, operations and essential physical services.

Scientific purpose
Research into industrial control resilience, instrumentation integrity, physical-process dependencies, and complex infrastructure failure modes.
Primary research questions
  • Which independently substantiated dependencies dominate loss of a designated physical service under permitted hypothetical disturbances?
  • Can a designated measurement be reconstructed from a trusted instrument state and unaltered acquisition lineage?
Typical researchers and partner laboratories
Critical-infrastructure laboratories, utilities, OT researchers, metrology institutes, process-safety groups and infrastructure consortia.
Related KRYOS capabilities
KRYOS Mission LockEvidence OSCORE-X Dependency ReconstructionCDAS Causal System ModelHypercube + Scenario Monte CarloHELIOS OT/Physics Council
Services
CYS-01 · CYS-02 · CYS-03 · CYS-04
View Services

02

Authorized Adversarial Cyber Operations Research

Defensive red-teaming and adversary-defender modeling under written authority.

Scientific purpose
Controlled defensive red-teaming, cyber-range simulation, adversary-defender modeling, detection engineering, and lawful cyber operations research.
Primary research questions
  • Which defensive detection, containment and architecture controls interrupt the most consequential hypothesized intrusion paths?
  • Does a defense retain required coverage when safe, synthetic test conditions vary beyond its training examples?
Typical researchers and partner laboratories
Defensive cyber laboratories, cyber ranges, research SOCs, authorized red and purple teams, and university security research groups.
Related KRYOS capabilities
Evidence OSMITRE ATT&CK MappingCORE-X System ReconstructionHypercube Graph SimulationHELIOS Defensive CouncilCRUCIBLE
Services
CYS-05 · CYS-06 · CYS-07 · CYS-08
View Services

03

AI Security and Autonomous Cyber Defense

Assurance for agentic systems, models and automated decisions.

Scientific purpose
Agentic AI assurance, adversarial machine learning, automated decision safety, and AI supply-chain security.
Primary research questions
  • Which input and tool conditions cause an agent to violate the defined authorization and trust-boundary contract?
  • Under what authorized distribution changes does the model cease to meet its declared reliability and security claims?
Typical researchers and partner laboratories
AI laboratories, ML security researchers, SOC automation researchers, AI assurance teams and model platform maintainers.
Related KRYOS capabilities
KRYOS-12FEvidence OSHELIOS Agent CouncilNIST AML TaxonomyHypercube Scenario BuilderCRUCIBLE
Services
CYS-09 · CYS-10 · CYS-11 · CYS-12
View Services

04

Quantum and Post-Quantum Security

Evidence-bounded cryptographic transition and quantum communication assurance.

Scientific purpose
Quantum communication security, cryptographic implementation assurance, post-quantum migration, and cryptographic dependency analysis.
Primary research questions
  • Which quantum-vulnerable cryptographic dependencies most critically affect a documented system and what migration sequence is supportable?
  • Does a claimed secure quantum channel meet its specified security assumptions under measured implementation conditions?
Typical researchers and partner laboratories
Quantum communication laboratories, physics groups, cryptography engineers, university research IT and enterprise security architects.
Related KRYOS capabilities
Evidence OSCORE-X Dependency GraphHELIOS Crypto CouncilHypercube Transition ScenariosARCS Standards OverlayCRUCIBLE
Services
CYS-13 · CYS-14 · CYS-15 · CYS-16
View Services

05

Scientific Computing and Research Security

Integrity of research data, software, laboratory networks and intellectual property.

Scientific purpose
Research data integrity, laboratory networks, scientific software assurance, and intellectual property protection.
Primary research questions
  • Which changes to scientific data or metadata would alter a material published or planned conclusion?
  • Which software dependencies and reproducibility failures could reverse a scientific result or security claim?
Typical researchers and partner laboratories
Universities, research laboratories, HPC institutes, scientific archives, lab IT officers and technology-transfer offices.
Related KRYOS capabilities
Evidence OSSIOSCORE-XHELIOS Research DataEvidence HypergraphCRUCIBLE
Services
CYS-17 · CYS-18 · CYS-19 · CYS-20
View Services

06

Cyber Conflict and Strategic Resilience

Scholarly, strategic and defensive analysis of conflict and continuity.

Scientific purpose
Cyber conflict escalation research, infrastructure resilience, defensive continuity planning, and security research prioritization.
Primary research questions
  • Under what explicitly modeled assumptions might a defensive cyber decision unintentionally raise escalation risk?
  • Which defensive investments and restoration sequences preserve the greatest essential service under constrained resources?
Typical researchers and partner laboratories
Cyber-policy institutes, defense research universities, infrastructure resilience researchers, research operations centers and independent review panels.
Related KRYOS capabilities
Evidence OSHELIOS Strategic CouncilCDAS Causal HypothesesGame TheoryHypercube Scenario TreesMonte Carlo if justified
Services
CYS-21 · CYS-22 · CYS-23 · CYS-24
View Services

Service catalog

24 KRYOS-Enabled Research Services

Each service is described from the institute's awardee manual. Deliverables are expected outputs, conditional on approved scope, real data, available tools, successful execution and release gates.

Showing 24 of 24 services

Architecture

The KRYOS Research Intelligence Architecture

KRYOS is a modular intelligence and adversarial-assurance architecture. Not every module is operational for every award, quantum hardware is not used automatically, and no universal superiority is claimed. Module availability is verified and recorded for each study. The complete catalog lives in the KRYOS framework library.
KRYOS v6 / Hypercube Intelligence

Modular orchestration layer that locks the mission, selects the minimum-sufficient framework stack and records which capabilities actually ran.

Evidence OS

Evidence-grounded retrieval and source manifests. Every proposition is classified by origin (observed, measured, simulated, inferred, asserted, unknown) and by maturity.

SIOS

Scientific system boundaries and charters, used where physics, quantum or measurement assumptions govern the research question.

CDAS

Causal system modeling that represents competing explanations and dependencies rather than a single preferred story.

HELIOS

Structured expert councils (for example OT, physics, defensive or program review) that examine models from discipline-specific perspectives.

OmniSynth

Synthesis of findings across sources and disciplines while keeping observation, inference and speculation separate.

CORE-X

System and dependency reconstruction from approved documentation, inventories and artifacts.

CRUCIBLE and Decisive Defect Search

Independent adversarial challenge of the most consequential assumption, evidence chain, benchmark or claimed improvement.

GAER

Defect repair and retest loop, run only when real tool access and test fixtures exist, with rerun receipts.

SINE

Signal and uncertainty integration used to report what remains unresolved and why.

ARCS and ARCF

Assurance gates for authorization, isolation, data rights, technical validity, safety and named human decision authority. A fatal open gate forces a hold.

Monte Carlo and uncertainty-aware simulation

Used only when justified, with disclosed distributions, seeds, sample counts, convergence and regime limits.

Digital twins and controlled scenario modeling

Bounded, isolated representations for defensive what-if analysis. Never connected to live operational systems.

Independent review and human-approved release

Qualified humans approve claim wording, information class, publication and any next action.

Capability run status, recorded for every tool

EXECUTED
Ran, with version, environment, inputs and run receipt.
AVAILABLE_NOT_USED
Available but not used for this study.
SPECIFIED_NOT_INTEGRATED
Designed, not yet integrated.
UNAVAILABLE
Not available for this award.
UNKNOWN
Status not established.

Awards

KRYOS Cybersecurity Research-Access Awards

Advanced research capabilities for qualified synergistic laboratories.

The James Scott Institute for Frontier Sciences offers selected cybersecurity research-access awards that provide approved researchers with access to specified KRYOS Hypercube capabilities and associated framework technologies. The awards are personally funded by James Scott and administered by the Institute for Critical Infrastructure Cybersecurity (ICIC).

Awards are non-cash. They do not provide financial assistance, stipends, reimbursements, equipment funding, unrestricted computing resources, or monetary grants.

Approved awardees receive the research-access rights and technical capabilities expressly defined in their written award scope, subject to authorization, technical availability, security requirements, and research-review conditions.

  1. 01

    Research Application

    The applicant identifies the scientific or defensive research problem, qualified laboratory, and relevant services.

  2. 02

    ICIC Eligibility and Authorization Review

    ICIC evaluates research alignment, legitimacy, qualifications, and applicable security boundaries.

  3. 03

    Technical and Scientific Review

    The institute evaluates whether the proposed research is feasible using the available KRYOS capabilities.

  4. 04

    Scoped Access Award

    Approved applicants receive a written, limited technology-access entitlement defining included modules, duration, conditions, and review requirements.

  5. 05

    Research Execution and Review

    The selected workflows proceed under approved conditions with evidence-based outputs, reproducibility requirements, and human review.

Eligibility

Who Should Apply

  • University cybersecurity and computer science laboratories
  • Physics and quantum research laboratories
  • National and government research laboratories
  • Research nonprofits and independent think tanks
  • Critical infrastructure research organizations
  • Advanced AI and autonomous systems laboratories
  • Defense-oriented academic research institutions
  • Scientific software, HPC, and instrumentation researchers

Qualified research alignment, legitimate project ownership, defensible methodology, and a specific technical question matter more than organizational size.

No organization category is automatically eligible. Each application is assessed for research legitimacy, authority over the systems and data involved, legal and safety eligibility, reviewer independence and available platform capacity.

Outputs

What Successful Awardees Receive

Depending on the selected service and actual capabilities, outputs may include the following. Not every awardee receives every output; deliverables depend on the approved award scope, available tools, successful execution and applicable release gates.
  • Evidence and technical research reports
  • Research assumption and uncertainty registers
  • Adversarial findings and verification matrices
  • Cyber-physical dependency models
  • Controlled simulation results
  • Scientific and security assurance assessments
  • Defensive engineering recommendations
  • Reproducible test specifications and artifacts
  • Baseline comparison and evaluation metrics
  • Conditional research recommendations
  • Human-reviewed technical decision packages

Boundaries

Research Integrity and Security Boundaries

Cyber conflict research at the institute is scholarly, strategic and defensive analysis. Every study passes non-compensatory release gates: a fatal open gate forces a hold, and no score offsets it.
  • Authorized research only.
  • Read-only assessment by default.
  • Isolated cyber-range or simulator use for active testing.
  • Partner control of research data.
  • Respect for scientific and institutional intellectual property.
  • No public submission of sensitive technical materials.
  • Independent scientific and security review when appropriate.
  • Explicit uncertainty and capability-status reporting.
  • No automatic authority for operational cyber actions.
  • No claims of government affiliation without verified authorization.
  • No DARPA relationship, endorsement or accreditation is implied.
  • Clear separation of research findings from strategic speculation.
G-AUTH
Written scope, data rights, asset ownership, cyber authorization and physical safety.
G-EVID
Source provenance, independence, contrary evidence and correct claim-origin labels.
G-CAP
Tool and framework execution evidenced; unimplemented functionality identified.
G-SCI
Physics, statistics, method assumptions and application regime are defensible.
G-EVAL
Baseline, task definition, estimator uncertainty and holdout validity are inspectable.
G-ADV
A competent independent challenger tests decisive defects and evaluator bias.
G-REPAIR
Material defects have owners, evidence needs, regressions and retest status.
G-RELEASE
A qualified human approves claim wording, information class, publication and next action.

Application

Apply for a Cybersecurity Research-Access Award

Three short steps. Share only non-sensitive scoping information. Do not include source code, credentials, exploit material, classified or export-controlled data, protected personal information, or detailed infrastructure attack surfaces.
  1. Step 1
  2. Step 2
  3. Step 3

Step 1. Applicant and organization