Market Synthocracy. A Definition and Typology of Synthocratic Market Power

From AI-assisted choice to the constitutional governance of economic participation

Abstract

Market synthocracy is a market order in which human beings and legally recognized institutions formally remain responsible for buying, selling, contracting, investing and allocating resources, while AI systems, autonomous agents, predictive models, protocols and digital platforms increasingly prepare, constrain and execute the conditions under which those economic decisions occur.

Market synthocracy does not require an artificial intelligence to become the legal owner of a company, the sovereign ruler of an economy or the formal signatory of every contract. It begins earlier. It emerges when synthetic systems determine:

which suppliers are visible;
which buyers are eligible;
which products can be understood;
which offers enter comparison;
how risks are classified;
which opportunities receive priority;
what conditions may be negotiated;
which transactions may proceed;
when human approval is required;
which decisions can be challenged;
whether participation without agentic infrastructure remains practically possible.

In Martin Novak’s contemporary use of the term, synthocracy is a decision order in which humans continue formally to govern, approve, choose and assume responsibility while AI and post-AI systems prepare the field in which those choices occur. Power consequently migrates from the final decision toward the upstream preparation of the decision environment: routing, ranking, classification, prediction, summarization, access and refusal.

This essay proposes a five-part typology of market synthocracy:

Soft market synthocracy — AI ranks, filters and recommends.
Operational market synthocracy — AI conducts commercial processes, negotiates and prepares transactions.
Transactional synthocracy — AI enters binding transactions within a delegated mandate.
Hard market synthocracy — meaningful market participation becomes practically dependent on agentic infrastructure.
Constitutional synthocracy — protocols, platforms and infrastructural operators determine the fundamental rights, permissions and standing of market participants.

The five types should not be understood only as successive stages. Soft, operational and transactional synthocracy primarily describe how much action is delegated. Hard and constitutional synthocracy describe how deeply the market itself has become dependent on, and governed through, synthetic infrastructure.

The decisive transformation therefore does not occur only when an agent is permitted to buy.

It occurs when the architecture surrounding buyers and sellers begins to determine who may enter the market, what they may do there and whether they can appeal, exit or take another route.

1. The conceptual foundation

According to the Novakian Paradigm Institute’s account of the concept’s origin, synthocracy in its present sense was initiated and developed by Martin Novak as part of a broader body of work on AI, legitimacy, institutional power, decision systems and the limits of machine authority. It was not introduced merely as a synonym for direct “rule by AI.” It was developed to describe the quieter condition in which AI participates in decision-making before institutions have adequately identified where practical power has moved.

The foundational definition is:

Synthocracy is a decision order in which humans formally continue to govern, approve, choose and take responsibility, while AI and post-AI systems increasingly prepare the field in which those choices occur.

This definition contains four essential elements.

1.1 Formal human continuity

Human beings remain visible within the institutional structure.

They may continue to:

approve budgets;
sign contracts;
select suppliers;
authorize payments;
accept liability;
occupy executive positions;
issue official explanations.

Synthocracy does not require the disappearance of these functions.

1.2 Synthetic preparation

AI systems increasingly perform upstream work such as:

filtering;
ranking;
classifying;
scoring;
summarizing;
predicting;
routing;
recommending;
generating defaults.

The human decision is made inside an environment already prepared by these operations.

1.3 Migration of practical power

The most consequential actor may no longer be the person performing the final visible act. It may be the system—or combination of systems—that determined which options became visible, which evidence was considered and which route reached the decision-maker.

Novak’s decision-environment concept identifies this field as the total combination of visible options, invisible filters, timing pressures, incentives, defaults, classifications, risk scores, reputational signals, institutional rules and machine-generated summaries through which a decision becomes possible.

1.4 Persistence of responsibility

The formal institution continues to hold a person responsible even where that person had limited control over the upstream preparation of the result.

This creates the possibility of the ceremonial human: an individual who remains at the final visible surface of the process because the human signature still carries legitimacy, even though the substance of agency has migrated elsewhere.

2. Definition of market synthocracy

Core definition

Market synthocracy is an economic decision order in which humans and legal institutions formally remain responsible for market participation and commercial commitments, while synthetic systems increasingly determine the visibility, admissibility, ranking, routing, negotiation, authorization and execution of economic opportunities and transactions.

A shorter definition is:

Market synthocracy arises when AI does not merely participate in the market but increasingly prepares and governs the conditions under which market participation becomes possible.

The definition has three dimensions.

2.1 Decision dimension

AI shapes or performs economic decisions concerning:

supplier selection;
pricing;
credit;
investment;
procurement;
logistics;
insurance;
employment;
inventory;
capacity allocation.

2.2 Execution dimension

AI can transform analysis into action by:

sending an RFQ;
making a counteroffer;
reserving stock;
issuing an order;
initiating payment;
scheduling delivery;
changing access conditions.

2.3 Infrastructural dimension

Participation becomes dependent on systems controlling:

identity;
product and company data;
discovery;
protocols;
registries;
reputation;
mandates;
payments;
dispute resolution.

Market synthocracy becomes structurally significant when all three dimensions begin to converge.

3. What market synthocracy is not

A precise definition requires clear boundaries.

3.1 It is not every use of software in commerce

A database that records inventory is not necessarily synthocratic.

A calculator used to determine a price is not automatically synthocratic.

A system becomes relevant to synthocracy when it affects:

who is considered;
which option is preferred;
how a participant is classified;
what opportunity is available;
what transaction may proceed;
who can challenge the outcome.

The Novakian distinction is between ordinary automation and automation that touches decisions, access, priority, risk, rights, duties, reputation or accountability.

3.2 It is not identical to algorithmic commerce

Algorithmic trading, dynamic pricing and recommendation engines may form parts of market synthocracy, but the concept is broader.

It includes the institutional order surrounding:

authority;
admissibility;
delegation;
identity;
routing;
human responsibility;
market access;
contestability;
infrastructural dependence.

3.3 It is not necessarily full autonomy

A system can be synthocratic even when a person approves every final transaction.

The relevant question is not only:

Did the system place the order?

It is also:

What had the system already determined before the person was asked to approve it?

3.4 It is not synonymous with monopoly

A competitive market containing many platforms can still be synthocratic.

Conversely, a concentrated market can remain largely human-directed.

Synthocracy describes the architecture of decision preparation and execution. Market concentration is a separate, though closely related, dimension.

3.5 It is not necessarily illegitimate

A synthocratic system can be:

authorized;
transparent;
contestable;
reversible;
pluralistic;
beneficial.

The term diagnoses the location of practical power. It does not, by itself, conclude that every such arrangement is unacceptable.

The normative question is whether synthetic power remains answerable to legitimate institutions and affected participants.

4. Why a typology is necessary

The statement “AI is used in the market” is analytically weak.

It does not distinguish between:

a product-ranking assistant;
an autonomous negotiating agent;
a system authorized to issue binding purchase orders;
a platform controlling supplier discovery;
a protocol defining which commercial actions can exist.

These arrangements distribute power differently.

A useful typology must therefore identify:

what the synthetic system does;
whether its output is binding;
where the human enters the process;
whether the participant can use an alternative route;
who controls the governing rules;
whether the outcome can be reconstructed and challenged.

The five types proposed here are defined by their dominant form of power:

Type	Dominant form of power
Soft market synthocracy	Epistemic and curatorial power
Operational market synthocracy	Procedural and negotiating power
Transactional synthocracy	Executable and committing power
Hard market synthocracy	Infrastructural and dependency power
Constitutional synthocracy	Rule-making and rights-defining power

5. Soft market synthocracy

Definition

Soft market synthocracy is a market condition in which AI systems filter, rank, score, summarize or recommend economic options while human beings retain formal authority to make and execute the final decision.

AI does not ordinarily enter a binding contract at this level.

It prepares the decision environment.

5.1 Typical functions

A soft market-synthocratic system may:

rank suppliers;
recommend products;
score leads;
classify customers;
identify risk;
compare quotations;
summarize contractual terms;
prioritize procurement cases;
recommend an award;
predict the probability of delivery failure;
determine which opportunities a user sees first.

5.2 Location of power

The central form of power is curatorial.

The system controls:

attention;
order;
salience;
classification;
relevance;
the candidate set.

It need not directly decide who wins. It can determine who is meaningfully seen.

A buyer may formally select a supplier, but the recommendation system may have excluded hundreds of possible suppliers before the buyer encountered the shortlist.

A manager may formally reject the recommended option, but doing so may require more time, information and institutional justification than accepting the default.

5.3 Human role

The human is usually described as:

the decision-maker;
the approver;
the reviewer;
the responsible manager.

That role is real only when the person can:

inspect the alternatives;
understand the criteria;
modify the ranking;
restore excluded options;
obtain underlying evidence;
override the recommendation without disproportionate penalty.

Without these capabilities, the human begins to become ceremonial.

The continued presence of a human signatory cannot, by itself, prove that human control remained intact.

5.4 Primary risks

Candidate-set invisibility

The system decides which participants enter consideration.

Automation bias

Users disproportionately accept system recommendations because they appear objective or data-driven.

Compression of alternatives

A complex market is reduced to a small number of machine-selected options.

Hidden value judgments

Weights assigned to price, risk, delivery, sustainability or reputation are presented as technical calculations rather than policy choices.

Responsibility displacement

The human carries formal responsibility for an outcome shaped upstream by systems they do not control.

5.5 Diagnostic test

A market is in soft synthocracy when the answer to the following question is yes:

Would a different ranking, classification, summary or recommendation system materially change the decisions that humans make, even though humans still formally approve them?

5.6 Commercial examples

Current enterprise procurement systems already illustrate parts of this level. Oracle describes agentic sourcing tools that monitor negotiations, surface exceptions and recommend award strategies, while SAP presents networks of procurement, supply-chain and risk agents that integrate information to support supplier-selection decisions.

These systems can provide substantial efficiency and analytical value. Their synthocratic significance lies in the fact that they can structure what procurement professionals perceive as the most rational choice.

6. Operational market synthocracy

Definition

Operational market synthocracy is a market condition in which AI systems autonomously conduct material parts of a commercial process—including discovery, communication, negotiation and transaction preparation—while final commitment generally remains subject to human or institutional approval.

The system moves beyond ranking.

It begins to act.

6.1 Typical functions

An operational agent may:

detect a purchasing need;
prepare an RFQ;
identify eligible suppliers;
invite them to participate;
request missing information;
normalize responses;
make or receive counteroffers;
negotiate price or delivery;
check internal policies;
prepare a contract;
recommend allocation;
draft a purchase order;
escalate exceptions.

6.2 Location of power

The dominant power is procedural.

The system determines:

how the process is structured;
which participant is contacted;
what questions are asked;
how many negotiation rounds occur;
which variables are negotiable;
when the process closes;
which exceptions reach a human.

This is more consequential than soft synthocracy because the system actively changes the commercial environment rather than merely describing it.

6.3 Human role

The human typically:

defines the strategy;
sets policy limits;
approves the final award;
manages exceptions;
reviews high-value cases.

The quality of this control depends on when the human enters.

A person who appears only after the system has:

selected the suppliers;
conducted the negotiation;
narrowed the alternatives;
prepared one recommended outcome,

may have formal veto power but limited practical authorship of the decision.

6.4 Current market direction

Commercial systems are already approaching this level.

Oracle’s Autonomous Sourcing Agent is designed to identify eligible low-value, high-volume requisitions, prepare sourcing events, invite suppliers and communicate according to organizational policies. Keelvar describes agents capable of receiving a request, constructing the sourcing event, engaging suppliers, negotiating, analysing bids and recommending an award. Pactum markets agents that execute negotiations autonomously or with buyer approval.

These examples do not prove the arrival of an autonomous universal market. They show that operational delegation is moving from conceptual prototypes into enterprise systems.

6.5 Primary risks

Procedural opacity

A company knows the result but cannot reconstruct the process that produced it.

Negotiation leakage

An agent may reveal urgency, budget tolerance or internal constraints.

Policy laundering

Organizational preferences encoded in the agent appear as neutral machine conclusions.

Automated exclusion

Suppliers are removed because of missing data, incompatible formats or misunderstood specifications.

Speed asymmetry

Participants with machine-readable systems respond immediately, while human-dependent suppliers miss the operational window.

Exception suppression

Cases that do not fit the dominant workflow are delayed, simplified or discarded.

6.6 Diagnostic test

Operational market synthocracy exists where:

Removing the agent would not merely reduce analytical convenience but would materially interrupt the commercial process itself.

7. Transactional synthocracy

Definition

Transactional synthocracy is a market condition in which an AI agent can create binding economic commitments, issue or accept orders, reserve resources or initiate payment within the boundaries of a prior human or institutional mandate.

At this level, the agent does not only prepare the transaction.

It commits the principal.

7.1 Typical functions

A transactional agent may:

accept a quotation;
place a purchase order;
confirm a sales order;
renew a subscription;
reserve inventory;
book transport;
purchase cloud resources;
modify contract quantities;
authorize a payment;
initiate settlement;
cancel or replace an order.

7.2 The mandate as the central institution

Transactional synthocracy depends on delegated authority.

A mandate may define:

maximum transaction value;
permitted products;
approved counterparties;
acceptable delivery dates;
geographical limits;
minimum documentation;
payment conditions;
duration;
escalation rules;
right of further delegation.

Google’s Agent Payments Protocol provides a current example of the movement toward mandate-based commerce. AP2 uses cryptographically signed digital mandates as evidence of user instructions and distinguishes between real-time approval and more detailed advance authorization for delegated transactions.

The mandate is necessary, but it does not resolve every question of legitimacy.

A signed instruction can prove that an agent received technical authorization. It does not automatically establish:

whether the mandate was understood;
whether it was proportionate;
whether its scope was appropriate;
whether the agent’s inferences were correct;
whether affected third parties could contest the outcome.

7.3 Capability versus authority

Transactional synthocracy makes Novak’s principle capability is not authority especially important.

A system may be capable of finding a lower price or concluding a transaction faster than a human. That capacity does not independently give the system the legitimate right to bind the company.

Capability asks whether the task can be performed effectively.

Authority asks who has the standing to decide and under what constraints.

Novak argues that accuracy, efficiency or statistical superiority cannot by themselves cross the boundary from ability to legitimate authority.

7.4 Human role

The human moves from transaction-level decision-maker to:

mandate designer;
policy owner;
limit setter;
exception reviewer;
audit authority.

This can strengthen human governance when mandates are precise and reviewable.

It can weaken it when people approve broad permissions they do not fully understand.

7.5 Primary risks

Mandate ambiguity

The agent’s formal authority is broader or less precise than intended.

Delegation chains

One agent delegates to another, making the true chain of authority difficult to reconstruct.

State divergence

The agent acts on inventory, price or credit data that have changed.

Irreversibility

Payment, production, logistics or resource reservation begin before a human detects an error.

Prompt-driven manipulation

External content affects product ranking, transaction selection or payment behaviour.

Research published in 2026 has identified potential governance gaps in contemporary agent protocols and experimental prompt-injection vulnerabilities in agent-payment implementations. These findings are early research rather than final judgments on all deployments, but they illustrate why technical authorization alone cannot substitute for broader governance.

7.6 Diagnostic test

Transactional synthocracy exists when:

An agent can create a legally, financially or operationally consequential commitment without a contemporaneous human decision, provided the action falls within a previously granted mandate.

8. Hard market synthocracy

Definition

Hard market synthocracy is a market condition in which meaningful participation becomes practically dependent on agentic infrastructure, synthetic classification and protocol compatibility, even where non-agentic alternatives continue formally to exist.

Hard synthocracy is not defined only by the autonomy of individual transactions.

It is defined by dependency.

8.1 The decisive change

In soft, operational and transactional synthocracy, the central question is:

How much of the commercial decision has been delegated?

In hard market synthocracy, the question becomes:

Can a firm still meaningfully participate without entering the synthetic infrastructure?

A supplier may remain legally free to send an email.

But if:

buyers accept only machine-readable RFQs;
agentic competitors respond in seconds;
inventory is reserved before the email is read;
compliance is verified through automated credentials;
payment depends on protocol-based authorization;
manual review takes days,

the manual route remains formally open but economically unusable.

8.2 Characteristics

Hard market synthocracy is present when several conditions converge.

Agent-determined visibility

Agents decide which businesses, products and offers enter consideration.

Protocol-dependent access

Participation requires compatibility with specified interfaces, schemas or credentials.

Infrastructural identity

A firm must possess a recognized machine-verifiable identity.

Synthetic reputation

Access, price and priority depend on machine-generated performance or risk scores.

Machine-speed competition

Human-only processes cannot meet the timing requirements of the market.

Platform dependency

Switching providers causes loss of:

reputation;
data;
transaction history;
customer access;
commercial identity.

Impractical manual fallback

Manual routes exist but carry severe delay, cost or service penalties.

Weak contestability

Participants cannot meaningfully challenge automated exclusion, routing or ranking.

8.3 The installation mechanism

Hard market synthocracy is unlikely to arrive through a formal declaration.

It will emerge through accumulated adoption.

Companies adopt agents for efficiency. Users adopt them for convenience. Governments and institutions adopt them for scale. Platforms adopt them for control and monetization. Each local choice appears rational, but their combined effect can produce a system on which everyone becomes dependent.

Novak describes this dynamic as synthocracy becoming a default operating system: not imposed through a coup, but installed through repeated decisions whose short-term incentives point in the same direction. Once decision-making becomes infrastructure, opting out can become a form of exclusion.

8.4 Market classes under hard synthocracy

Hard market synthocracy can produce new commercial classes.

Agent-native participants

Their data, identities, products and processes are directly executable.

Agent-compatible participants

They can interact with agentic infrastructure but still depend on some human processes.

Gateway-dependent participants

They require an intermediary to translate documents, emails and human knowledge into protocol-compatible transactions.

Human-exception participants

They enter the market only through costly manual review.

Machine-invisible participants

They legally exist but do not appear in the executable market perceived by agents.

Blocked participants

They fail identity, risk, compliance or protocol requirements and cannot access dominant commercial networks.

8.5 Primary risks

Market foreclosure without formal exclusion

A firm disappears because it cannot satisfy technical participation conditions.

Dependency without consent

Organizations become reliant on systems they adopted incrementally rather than deliberately.

Synthetic path dependence

Early rankings create more transactions, which create better performance data, which produce still higher rankings.

Infrastructural lock-in

The cost of leaving a platform becomes greater than the cost of accepting its governance.

Right without remedy

Participants formally possess legal or contractual rights but cannot exercise them at machine speed.

8.6 Diagnostic test

Hard market synthocracy exists where:

A participant retains the formal right to trade without the dominant infrastructure but loses the practical ability to compete, obtain access or complete transactions by doing so.

9. Constitutional synthocracy

Definition

Constitutional synthocracy is a market order in which protocols, platforms and infrastructural operators determine the fundamental standing, rights, permissions, obligations and remedies of economic participants.

Constitutional synthocracy is the deepest form of market synthocracy because it concerns not only decisions or transactions, but the rules determining what kinds of decisions and transactions can exist.

9.1 Why “constitutional”?

A constitution establishes the fundamental structure of an order.

It defines:

who is recognized;
who has authority;
which powers exist;
how powers are limited;
which rights participants possess;
how rules are changed;
how disputes are resolved;
whether exit or appeal is possible.

Agentic market infrastructures increasingly perform analogous functions.

They may determine:

what constitutes a valid commercial identity;
what counts as an admissible product;
which agents may communicate;
which actions require mandates;
what constitutes a valid offer;
how reputation is calculated;
which participants receive access;
what evidence is accepted;
when a transaction becomes final;
how an exclusion can be appealed.

9.2 Protocols as constitutional instruments

Protocols such as A2A, UCP and AP2 are currently being developed as open technical foundations for agent communication, commerce and mandate-based payments. A2A provides a common language for interoperable agents and uses Agent Cards to describe identity, capabilities, endpoints and authentication requirements. UCP is intended to provide shared commercial building blocks from discovery through checkout. AP2 provides a structure for verifiable transactional mandates.

These protocols do not independently constitute an entire market constitution.

Their significance lies in the institutional functions that can be built around them:

registries;
identity providers;
certification systems;
reputation layers;
platform-specific extensions;
payment rules;
access policies.

A technically open protocol can coexist with a commercially closed implementation.

9.3 Layer C and constitutional admissibility

Martin Novak’s concept of Layer C provides a useful framework for understanding constitutional market power.

Layer C is the pre-runtime admissibility layer: the threshold at which a system determines whether a state, claim, signal or decision may enter the field where execution becomes possible. It precedes runtime rules and asks not only how an admitted action should be governed, but whether it should have entered the executable field at all.

Applied to markets, the constitutional questions are:

Which firms have the right to enter the supplier registry?
Which product categories are recognizable?
Which credentials count as evidence?
Which agents are trusted?
Which transactions are admissible?
Which participants are routed to review?
Which entities remain outside execution entirely?

This is more fundamental than deciding who wins a transaction.

It defines who and what can become a candidate for transaction.

9.4 Constitutional power holders

The constitutional administrators of an agentic market may include:

protocol foundations;
platform operators;
cloud providers;
model providers;
identity providers;
payment networks;
certification authorities;
risk-data vendors;
enterprise software providers;
regulators.

No single institution needs to control the entire architecture.

Constitutional power can emerge from the interaction of multiple layers.

9.5 Private constitutionalism

A particularly important case occurs when private platforms determine fundamental economic rights through:

terms of service;
API policies;
ranking systems;
identity requirements;
commercial certification;
proprietary protocol extensions.

Digital-market regulation already recognizes that very large platforms may operate as gatekeepers between businesses and users. The EU Digital Markets Act imposes obligations on designated gatekeepers because their platform position can create structural bottlenecks in digital markets.

Agentic markets could extend this gatekeeping power beyond visibility and distribution into delegated negotiation and execution.

9.6 Constitutional synthocracy can be open or closed

The term is not necessarily synonymous with dystopia.

Accountable constitutional synthocracy

The infrastructure embeds:

transparent rules;
portable identity;
multiple registries;
rights of appeal;
auditable mandates;
effective human review;
non-discriminatory access;
interoperable protocols;
functional fallback routes.

Unaccountable constitutional synthocracy

The infrastructure embeds:

opaque access criteria;
non-portable reputation;
proprietary extensions;
concentrated identity control;
unchallengeable ranking;
ceremonial human oversight;
no effective exit;
no practical manual alternative.

The essential question is not whether a market has a synthetic constitution.

A complex agentic market will almost certainly require one.

The question is:

Who writes it, whose interests it serves, who can challenge it and how it can be changed.

9.7 Diagnostic test

Constitutional synthocracy exists where:

Changing technical standards, platform policies or infrastructural rules alters the fundamental rights and market standing of participants without requiring a corresponding change in public law or individually negotiated contracts.

10. The five types are not a simple ladder

It would be misleading to present the typology as:

Soft → Operational → Transactional → Hard → Constitutional

The first three types concern the degree of delegated action.

The final two concern the depth of infrastructural governance.

A more accurate model uses two axes.

Axis A: delegation of commercial action

Soft
→ Operational
→ Transactional

This axis asks:

How much of the decision and transaction has moved from direct human action to synthetic systems?

Axis B: dependence and rule-making

Optional infrastructure
→ Hard dependence
→ Constitutional governance

This axis asks:

How deeply does synthetic infrastructure determine participation, rights and remedies?

A market can therefore be:

soft but constitutional;
transactional but not hard;
hard but only partly transactional;
operational and constitutionally open;
transactional and constitutionally closed.

Example: soft but constitutional

A dominant marketplace does not permit autonomous purchases, but its ranking system and access rules determine which sellers are visible and which can reach customers.

AI remains advisory at the transaction level.

The platform is already constitutional at the market-access level.

Example: transactional but not hard

A company permits an internal agent to replenish inventory automatically from a small group of approved suppliers.

The agent enters binding transactions.

Suppliers can still trade through other channels, and the wider market is not dependent on this infrastructure.

Example: hard but not fully transactional

Most buyers require machine-readable supplier profiles and automated compliance verification, but humans still approve every order.

Market participation is infrastructurally dependent even though transaction-level autonomy remains limited.

This distinction is crucial.

The most serious form of market power may appear before full transactional autonomy.

11. Comparative typology

Dimension	Soft	Operational	Transactional	Hard	Constitutional
AI ranks and recommends	Central	Central	Present	Present	May be present
AI communicates externally	Limited	Central	Central	Usually required	Governed by protocol
AI negotiates	Usually no	Yes	Yes	Often	Rules define permission
AI makes binding commitment	No	Normally no	Yes, within mandate	Often expected	Defined structurally
Human approves each transaction	Usually	Usually	Not always	May become impractical	Depends on constitution
Market participation requires agentic infrastructure	No	No	Not necessarily	Yes in practice	Usually institutionalized
Protocols define rights and standing	Limited	Growing	Important	Structurally important	Fundamental
Manual fallback	Strong	Available	Available	Weak or penalized	Constitutionally defined or absent
Primary power	Curatorial	Procedural	Executable	Infrastructural	Constitutive
Main danger	Hidden shaping	Process capture	Unauthorized commitment	Dependency and exclusion	Private rule without accountability

12. Threshold indicators

A market moves from one type to another when specific thresholds are crossed.

Soft-to-operational threshold

Crossed when AI begins to:

initiate commercial communication;
construct processes;
request data;
make proposals;
manage negotiation steps.

Operational-to-transactional threshold

Crossed when AI can create an external commitment without contemporaneous human approval.

Optional-to-hard threshold

Crossed when refusing the synthetic path causes material exclusion from:

customers;
suppliers;
prices;
payment;
timing;
service quality.

Hard-to-constitutional threshold

Crossed when infrastructural rules determine enduring rights and market standing, including:

identity;
admissibility;
discoverability;
portability;
appeal;
finality.

13. The market-synthocracy assessment framework

An institution can evaluate its market architecture through ten questions.

13.1 Visibility

Who determines which counterparties or products become visible?

13.2 Admissibility

What conditions must be satisfied before a participant enters the executable process?

13.3 Classification

Which system defines categories, risks and equivalence?

13.4 Authority

Under whose mandate does each agent act?

13.5 Negotiation

Which variables can an agent change, and within what boundaries?

13.6 Commitment

Can the system create binding financial or contractual obligations?

13.7 Routing

Who determines whether a case reaches a human, another agent, a delay or a refusal?

Novak’s work on routing emphasizes that a right to speak or appeal can become ineffective if the system routes a participant outside the zone where the intervention produces consequences.

13.8 Contestability

Can an affected participant obtain reasons, correct data and request a different route?

13.9 Fallback

Does a real, timely and economically usable non-agentic route remain available?

13.10 Administration

Who can change the rules, revoke identity, modify access or terminate the infrastructure?

The answers reveal not only the level of automation, but the location of market power.

14. The role of human authority

Market synthocracy should not be evaluated through the binary distinction:

human decision / machine decision

A more useful model distinguishes several human roles.

Human as originator

The person initiates and defines the transaction.

Human as policy designer

The person establishes the agent’s mandate and limits.

Human as supervisor

The person monitors action and intervenes when necessary.

Human as exception handler

The person deals only with cases the system cannot resolve.

Human as approver

The person accepts or rejects the prepared outcome.

Human as signatory

The person gives formal legal effect to an outcome already determined in practice.

Human as liability holder

The person or institution remains responsible even when meaningful control is absent.

The lower the human moves in this sequence, the more likely formal authority has separated from practical agency.

15. The normative divide: accountable and unaccountable market synthocracy

The central objective should not be to preserve every manual commercial process.

Nor should it be to prohibit agentic transactions.

The relevant distinction is between accountable and unaccountable market synthocracy.

Accountable market synthocracy

It includes:

explicit mandates;
bounded autonomy;
portable identities;
traceable delegation;
meaningful reasons;
contestable rankings;
correction rights;
independent audit;
operational stoppability;
human intervention before irreversible commitment;
functional alternative routes.

Unaccountable market synthocracy

It includes:

hidden candidate selection;
opaque scoring;
unclear authority;
uncontrolled delegation;
non-portable reputation;
ceremonial human approval;
irreversible machine-speed execution;
ineffective appeal;
infrastructural lock-in;
no identifiable administrator with responsibility.

The difference cannot be reduced to the accuracy of the AI.

A highly accurate system can remain illegitimate if it lacks authority, answerability or contestability.

Conversely, a less sophisticated system may be institutionally acceptable when its role is transparent, bounded and reversible.

16. Market synthocracy as a field of research

The typology creates several distinct research programmes.

Soft market synthocracy research

Study:

recommendation;
ranking;
candidate-set formation;
synthetic visibility;
decision compression;
automation bias.

Operational market synthocracy research

Study:

agentic procurement;
autonomous negotiation;
synthetic workflow governance;
exception routing;
human oversight.

Transactional synthocracy research

Study:

mandates;
delegation;
contractual agency;
payment authorization;
reversibility;
liability.

Hard market synthocracy research

Study:

infrastructural dependence;
machine-readable market access;
manual fallback;
MŚP exclusion;
gateway dependency;
market concentration.

Constitutional synthocracy research

Study:

protocol governance;
platform rule-making;
identity infrastructure;
commercial rights;
registry governance;
portability;
appeals;
digital-market constitutionalism.

17. Formal definition

A market may be described as synthocratic to the degree that the following conditions hold:

[
MS = f(P, D, E, I, C, A)
]

Where:

(P) = synthetic preparation of the decision environment;
(D) = delegation of commercial action;
(E) = autonomous execution;
(I) = infrastructural dependence;
(C) = constitutional control over participation;
(A) = accountability deficit between practical control and formal responsibility.

The purpose of the formula is conceptual rather than computational.

It shows that market synthocracy is multidimensional.

A market may score highly in synthetic preparation but weakly in autonomous execution. Another may permit extensive autonomous execution while preserving open infrastructure and meaningful contestability.

The most dangerous configuration is not simply maximum automation.

It is:

[
High\ P + High\ D + High\ E + High\ I + High\ C + High\ A
]

That is a market in which synthetic systems prepare decisions, act, commit, control access and define rights while responsibility remains attached to humans or institutions unable to reconstruct or redirect the process.

Conclusion

Market synthocracy names the transition from AI as a commercial tool to AI-mediated infrastructure as a governing economic environment.

Its earliest form is soft and curatorial.

AI ranks, filters and recommends while humans formally choose.

Its operational form turns recommendation into process.

Agents discover, communicate, negotiate and prepare transactions.

Its transactional form turns process into commitment.

Agents bind organizations within the limits of delegated mandates.

Its hard form turns convenience into dependency.

Participation becomes practically impossible without synthetic identity, machine-readable data, protocols and platforms.

Its constitutional form turns infrastructure into an economic order.

Protocols and platform operators define who exists, who may enter, what actions are permitted, which evidence counts, how decisions become final and whether an affected participant may appeal or exit.

The central insight follows directly from Novak’s theory of the decision environment:

Market power no longer resides only in the actor who makes the final decision. It increasingly resides in the systems that prepare what can be found, admitted, negotiated and executed.

Market synthocracy therefore does not begin when an AI becomes the visible ruler of commerce.

It begins when the human still chooses, the manager still approves and the company still signs—but the field in which these acts occur has already been synthetically prepared.

The central question is no longer merely:

Who made the transaction?

It is:

Who governed the conditions under which that transaction—and no other—became possible?