MGXAI ®
Introducing  /  MGXAI Mechanistic Systems™

Mechanistic systems reverse engineering across computational, automated, and AI enabled systems

Evidence supported reconstruction. Verifiable system records. Continuity as authorized components change.

AI increasingly participates in discovery, recommendations, customer interactions, workflows, routing, decisions, and commercial outcomes. But the complete journey behind an outcome may be distributed across several independent systems, and each participant may preserve only the portion visible from its own system.

MGXAI Mechanistic Systems™ is designed to reconstruct an evidence supported account of these distributed journeys and preserve a verifiable system record of what the available evidence supports occurred.

Understand what happened. Trace what influenced it. Verify what the evidence supports. Preserve continuity as systems change.

The AI mediated journey before and after the MGXAI ® external structure Six independent systems are shown in sequence: creator, AI assistant, application, referral system, merchant, and conversion. Each holds only a partial view of the journey. When the MGXAI ® external structure is shown, a surrounding frame and connecting evidence paths appear, making the relationships, sequence, provenance, attribution, controls, and evidence examinable together. MGXAI ®® EXTERNAL STRUCTURE CREATORAI ASSISTANTAPPLICATION REFERRALMERCHANT CONVERSION NO SINGLE SYSTEM HOLDS THE COMPLETE JOURNEY

Separate systems. Fragmented records. No unified journey.

01

AI mediated journeys are fragmented

A traditional software application usually understands the activity occurring inside its own boundaries. An AI mediated journey is different. It may move across systems that do not share the same records, rules, identifiers, or understanding of what happened.

Consider a commercial discovery journey. A creator publishes content. A person discovers a product. An AI assistant later encounters information about that product and summarizes or recommends it. The person interacts with another service. A referral or tracking event occurs. The customer completes a purchase. The merchant records the conversion.

The gap The final transaction exists. The complete path that influenced that transaction may not exist as one continuous record.

Important questions become difficult to answer

Each question below points at a different part of the journey. Together they describe what an organization may need to establish after an outcome has already occurred.

ORIGINATION

Where did the interest originate?

SEQUENCE

What happened after discovery?

PARTICIPANTS

Which systems participated?

MEDIATION

Did an AI system influence the journey?

STATE CHANGE

Did software alter the referral path?

ATTRIBUTION

What system ultimately received attribution?

EVIDENCE

Was the final attribution supported by evidence?

PERMISSIONS

Were required permissions respected?

CONTROLS

Did organizational controls remain in place?

HUMAN AUTHORITY

Was human review required?

REVIEWABILITY

Can another authorized party examine the evidence?

02

Reconstructing the mechanism behind the outcome

Mechanistic Systems examines the available evidence surrounding an AI mediated outcome and reconstructs an evidence supported account of the systems and events that contributed to it.

ARCHITECTURECOMPONENTSINTERACTIONS DEPENDENCIESSEQUENCESSTATE CHANGES PERMISSIONSAUTHORITYPROVENANCE ORIGINATIONMEDIATIONATTRIBUTION CONTROLSGOVERNANCEHUMAN INVOLVEMENT SYSTEM CHANGESAVAILABLE EVIDENCE

The goal is not to fill gaps with assumptions. The goal is to establish what the available evidence can support.

MGXAI ® distinguishes between supported findings, areas requiring further review, and what remains unresolved.

MGXAI ® distinguishes between what the evidence supports, what requires further review, and what remains unresolved. If part of a journey cannot be established from available evidence, the system should not represent an unsupported assumption as fact.

Evidence discipline The question is not: Can we create a convincing explanation? The question is: What can the available evidence actually support?
03

The exoskeleton

An AI mediated journey may cross several independent systems. No single participating system necessarily preserves the entire journey. MGXAI ® provides an external structure around the journey so important activity can remain examinable.

The MGXAI ® evidentiary exoskeleton The AI mediated journey sits at the center as a sphere. Concentric structural rings surround it carrying the layers architecture, interactions, sequences, state, control flow, provenance, origination, mediation, attribution, authority, controls, continuity, evidence, governance, human oversight, and delivery. Selecting a layer leans the central sphere toward that layer and reveals a short explanation of the question it helps examine. ARCHITECTURE INTERACTIONS SEQUENCES STATE CONTROL FLOW PROVENANCE ORIGINATION MEDIATION ATTRIBUTION AUTHORITY CONTROLS CONTINUITY EVIDENCE GOVERNANCE HUMAN OVERSIGHT DELIVERY THE AI MEDIATED JOURNEY CONVERSION
Select a layer

Seventeen structural layers

Each layer holds a different edge of the same reconstruction. Choose one to see the question it helps examine.

The exoskeleton principle Systems change. Evidence should survive the change.

MGXAI ® is not attempting to become every system represented in the journey. MGXAI ® surrounds the journey with an examinable structure.

Function one

Evidentiary structure

An AI interaction may be temporary. A model response may change. A browser session may end. A referral state may change. A workflow may continue. A customer may complete an outcome later.

Mechanistic Systems is designed to preserve enough reviewable structure around important events so organizations can later examine the available evidence surrounding the outcome.

Understand what happened.
Function two

Operational structure

A computational component may also change. A model can become unavailable. Access can change. A provider can change availability. An authorized user may no longer have access to a particular model.

The surrounding organizational function should not necessarily disappear because one computational component changed.

Preserve continuity as systems change.

Mechanistic Systems can support examination of what already occurred and preservation of continuity as systems change.

04

Five capabilities define the MGXAI ® perimeter

Each capability answers a different question about an AI mediated outcome. Select one to expand it.

01 RECONSTRUCT Understand what systems participated and what occurred
  • Which components participated
  • How components interacted
  • What operational dependencies existed
  • What relevant events occurred
  • In what order the evidence supports those events occurring
  • What state changed
  • What happened before and after a significant action
  • Where available records contain gaps

Missing evidence may be significant. Missing evidence is not proof that an event occurred.

02 TRACE Follow provenance, origination, mediation, and attribution

Modern discovery does not always move directly from advertisement to click to purchase. A person may first encounter an idea through a creator. Later, an AI assistant may summarize or recommend the same product. Another system may mediate the journey. A merchant may ultimately record the conversion.

  • Where discovery began
  • What mediated the interaction
  • Which systems participated before conversion
  • How attribution changed between origination and outcome
03 VERIFY Examine capability, action, authority, and control

An AI system producing an action does not automatically mean the action was authorized or properly governed. Four separate questions matter.

Capability

What was the system technically capable of doing?

Action

What does the available evidence support actually occurred?

Authority

Was the actor or component authorized to perform the action?

Control

Were required controls present when expected?

Three questions, in order. What was technically possible? What does the evidence show actually occurred? Was the actor or component authorized to do it?

Capability. Action. Authority. Evidence. Technical capability and organizational authority are not the same thing.

04 SUSTAIN Model portability without losing system continuity

An AI native system should not necessarily be inseparable from one computational component. When an approved component becomes unavailable, MGXAI ® can support substitution of another authorized component while preserving the surrounding business function and relevant evidence.

  • Reduced dependency on a single model
  • Greater operational resilience
  • A clearer record of component changes
  • Preservation of relevant controls
  • Continued human oversight
  • Reviewability of what changed

Substituting an authorized computational component while preserving the surrounding workflow, relevant controls, provenance, and reviewable evidence of what changed.

The component can change without unnecessarily losing system continuity or evidentiary history.

05 PROVE The final result should be reviewable

The output of Mechanistic Systems is intended to be a verifiable system record: a structured account of material findings in which the evidence supporting those findings remains available for authorized review.

A verifiable system record is a structured record of the reconstructed journey in which material findings are connected to the evidence supporting them, allowing an authorized second party to independently review the basis for the reconstruction.

  • Audit review
  • Attribution disputes
  • Compliance investigations
  • AI governance review
  • System migrations
  • Incident reconstruction

A narrative asks the reviewer to trust the storyteller. A verifiable record gives the reviewer evidence to examine.

05

Attribution integrity

Who received credit is only the first question.

Traditional attribution systems are generally designed to determine which source receives credit for a conversion according to the rules and signals available to that system. MGXAI ® asks a different and additional question: does the available evidence support how that credit was assigned?

Attribution journey from creator to conversion A journey moves through creator, discovery, AI mediation, software interaction, referral change, merchant, and conversion. A traditional tracking system reads final credit at the last recorded event. MGXAI ® expands the view to the broader evidence journey, where supported origination is examined separately from final credit. Supported origination and final credit are not necessarily the same party. CREATORDISCOVERYAI MEDIATION SOFTWAREREFERRAL CHANGEMERCHANTCONVERSION FINAL CREDIT SUPPORTED ORIGINATION EXAMINED ACROSS THE FULL JOURNEY SUPPORTED ORIGINATION AND FINAL CREDIT ARE NOT NECESSARILY THE SAME PARTY
STAGE 1 OF 8
Stage 1

A creator publishes content and creates discovery.

MGXAI ® does not begin with who should receive credit. It examines what the evidence supports.

Attribution tracking asks

Who received credit?

Attribution integrity asks

How was that credit assigned, what influenced it, and does the evidence support that it was validly earned?

AI mediated discovery can separate the source that created interest from the system that ultimately records the conversion. Mechanistic Systems reconstructs the available referral, discovery, mediation, and conversion evidence so organizations can examine how credit changed and whether the final attribution is supported.

MGXAI ® does not declare that the first creator always deserves the credit. MGXAI ® examines the evidence behind how credit changed. Tracking sees an attribution event. MGXAI ® examines the evidence surrounding the attribution event.

06

Continuity when a component changes

Organizations are increasingly building important workflows around external AI models. That creates a dependency. The organization may control the workflow but not the continued availability of every model participating in it.

Controlled substitution of an approved computational component An organizational framework containing workflow, rules, authority, human approval, evidence, and governance surrounds a computational component position. When Model A becomes unavailable the surrounding structural layer remains visible, and an approved alternative, Model B, enters the same controlled position. System continuity and evidence are preserved. ORGANIZATIONAL FRAMEWORK WORKFLOWRULESAUTHORITY HUMAN APPROVALEVIDENCEGOVERNANCE MODEL A ACTIVE MODEL B APPROVED ALTERNATIVE

An approved computational component operates inside the organizational framework.

When an approved component becomes unavailable, MGXAI ® can support substitution of another authorized computational component while preserving the surrounding business function and the evidence surrounding the change.

This capability is never a means of circumventing model restrictions, provider controls, licensing requirements, security controls, or user permissions. MGXAI ® supports authorized alternatives. It does not bypass unauthorized access restrictions.

07

Why verifiable evidence matters

Trust should not depend only on the final answer.

AI systems increasingly participate in decisions and workflows that affect businesses, creators, customers, partners, operations, and financial outcomes. As their role increases, organizations may need more than an answer. They may need evidence explaining the basis for important outcomes.

Possible applications

Internal review

Audit review

Attribution disputes

Partner disputes

Compliance investigation

AI governance review

System change review

Operational incident reconstruction

Human oversight review

Commercial reconciliation

Model substitution review

Vendor transition review

The purpose is not to promise that every question can always be answered.

The quality and depth of reconstruction depend on the evidence available to the organization and participating systems.

Evidence standard The purpose is to make supported findings reviewable and unsupported conclusions identifiable.
The larger point It is not simply about adding AI to an organization. It is about helping the organization continue operating as AI technology changes while preserving the evidence, attribution integrity, governance, and human accountability needed to understand important outcomes.
Evidence and Continuity

Evidence integrity and authorized system continuity

Evidence bounded reconstruction

MGXAI Mechanistic Systems™ reconstructs computational and automated system activity from available technical and organizational evidence. The integrity of the reconstruction depends on distinguishing supported findings, unresolved findings, and evidentiary gaps rather than representing unsupported assumptions as established events.

Not every historical event can always be reconstructed. The depth of reconstruction depends on the evidence available from the systems, components, records, and authorized sources participating in the analysis.

Authorized system continuity

MGXAI Mechanistic Systems™ supports integration, reconfiguration, migration, and substitution of authorized computational components while preserving relevant technical controls, provenance, human oversight, system continuity, and reviewable records of what changed.

Authorized continuity does not bypass provider restrictions, access controls, licensing requirements, security controls, or user permissions. Computational alternatives remain subject to the authorities and controls applicable to the surrounding system.

Services

What MGXAI Mechanistic Systems™ Provides

MGXAI Mechanistic Systems™ provides technological research in mechanistic systems, computer system analysis, and mechanistic systems reverse engineering across computational and automated systems, including systems containing deterministic and AI enabled components. Its online technology services support evidence based reconstruction of system architectures, component interactions, operational dependencies, state transitions, control flows, event sequences, provenance, technical controls, and computational outcomes.

MGXAI Mechanistic Systems is provided through online non downloadable software and related technological services for mechanistic systems analysis, mechanistic systems reconstruction, computational provenance analysis, reconstruction of system behavior, and generation and preservation of verifiable computational system records.

Where artificial intelligence is used, MGXAI provides AI enabled services as part of the broader Mechanistic Systems architecture rather than treating AI as the entire system.

Mechanistic Systems Research

MGXAI conducts technological research and analysis concerning how complex computational and automated systems can be reconstructed, examined, governed, reconfigured, and supported by verifiable computational evidence. Research may examine deterministic components, artificial intelligence enabled components, their interactions, system state changes, provenance, authority, technical controls, and the evidence available to reconstruct system behavior.

Mechanistic Systems Reverse Engineering

MGXAI Mechanistic Systems performs mechanistic systems reverse engineering across computational and automated systems. The service reconstructs and analyzes available technical evidence to identify system architecture, participating components, component interactions, operational dependencies, control flows, state transitions, event sequences, system changes, permissions, authority, provenance, and technical controls.

Mechanistic systems reverse engineering may examine systems composed of deterministic software, AI enabled components, or combinations of both. The purpose is to determine what the available evidence supports occurred, how system components interacted, what changed, and how relevant computational events contributed to an outcome.

10

Who Mechanistic Systems is for

Primary commercial applications

Merchants

Understand AI influenced commercial journeys, referral changes, attribution outcomes, and evidence supporting a conversion path.

Affiliate networks

Examine how commercial credit changed across software environments and whether available evidence supports the attribution outcome.

Agencies

Give clients a stronger basis for reviewing AI mediated campaigns, creator influence, automated recommendations, and commercial outcomes.

Creator platforms

Help preserve evidence of origination and influence when discovery begins with a creator but conversion occurs elsewhere.

Broader AI systems applications

AI native businesses

Design systems with continuity, reviewability, governance, and evidence preservation in mind.

Compliance and governance teams

Review whether technical actions remained within expected permissions, controls, and human authority requirements.

Audit and risk professionals

Obtain a structured basis for examining important AI assisted actions and evidence supporting material findings.

Technology leadership

Reduce unnecessary dependency on a single computational component while improving the ability to examine changes across the surrounding system.

11

Questions Mechanistic Systems is designed to help examine

  • What systems participated?
  • What happened?
  • What happened first?
  • What changed?
  • What information influenced the outcome?
The outcome standard While individuals expect AI to produce a result. MGXAI ® helps preserve the evidence needed to understand and verify how the result occurred.
13  /  Early access pricing

Priced against the decision it defends

MGXAI ® sells verifiable evidence, not software seats. The unit of value is a reviewable record that survives a dispute, an audit, or a partner negotiation.

Seats measure headcount. Value scales with journeys examined, receipts issued, and disputed dollars defended. Pricing below reflects early access terms while the validation programme is open.

Evidence supply

Creator

Free to start
Creator Console $29 monthly
  • Register your content and hold your own origination evidence
  • Generate referral carriers for the work you publish
  • Console: Creator Earnings Statements as signed documents
  • Console: proof of origination exports
  • Console: dispute packet generation when credit is contested
Join the creator registry

Creators are the origination layer of AI mediated commerce. Registration stays free by design.

MOST COMMON START Merchants, agencies, growth networks

Professional

$6,000 monthly
30,000 receipts included
  • Unlimited merchant programs
  • Settlement Statement generation for merchant and creator payouts
  • Dispute case support with evidence packets
  • Overage metered at 25 cents per receipt beyond allotment
Get started
Networks, large merchants, platforms

Enterprise

Engagement priced
Begins with a fixed scope audit
  • Attribution integrity audit of your network or merchant data exports, delivered as findings with evidence commitments
  • Audit fee credits toward an annual platform contract on conversion
  • Annual platform fee plus metered usage per verifiable receipt
  • Ongoing integrity monitoring across partners and delivery surfaces
  • Receipt licensing for networks that resell or embed MGXAI ® verification
Request the enterprise briefing

Invoiced with ACH or wire terms. Scoped to the systems, volume, and evidence involved.

Receipts, not seats

A seat measures who logged in. A receipt measures a commercial decision that can now be defended. Revenue maps to the value delivered, not to headcount.

Audits first

Enterprise relationships begin with a fixed scope audit: a defined engagement, a defined deliverable, findings your team can act on.

Creators stay free

Creators are the evidence supply side of AI mediated commerce, not the revenue centre. Registration and carrier generation stay free so the origination registry keeps growing.

References

Industry and standards context

The following references provide professional context for concepts such as AI risk management, accountability, traceability, provenance, and affiliate relationships. They do not validate MGXAI ® or its technical performance.

National Institute of Standards and Technology

Elham Tabassi. Artificial Intelligence Risk Management Framework, AI RMF 1.0. NIST AI 100 1. National Institute of Standards and Technology, 2023.

https://doi.org/10.6028/NIST.AI.100-1
National Institute of Standards and Technology

Chloe Autio, Reva Schwartz, Jesse Dunietz, Shomik Jain, Martin Stanley, Elham Tabassi, Patrick Hall, and Kamie Roberts. Artificial Intelligence Risk Management Framework: Generative Artificial Intelligence Profile. NIST AI 600 1. National Institute of Standards and Technology, 2024.

https://doi.org/10.6028/NIST.AI.600-1
World Wide Web Consortium

Timothy Lebo, Satya Sahoo, and Deborah McGuinness, editors. PROV O: The PROV Ontology. W3C Recommendation, 2013.

https://www.w3.org/TR/prov-o/
Organisation for Economic Co operation and Development

OECD AI Principles.

https://www.oecd.org/en/topics/ai-principles.html
Federal Trade Commission

FTC Endorsement Guides: What People Are Asking.

https://www.ftc.gov/business-guidance/resources/ftcs-endorsement-guides-what-people-are-asking

External references are provided for informational, standards, regulatory, and industry context only.

Reference to NIST, W3C, OECD, the Federal Trade Commission, or another external organization does not imply endorsement, certification, affiliation, approval, or validation of MGXAI ® or MGXAI Mechanistic Systems™.

MGXAI ® capabilities may vary according to implementation, authorized system access, evidence availability, organizational configuration, and the systems participating in a particular journey.