AIFC
Back to version

Published version: AIFC-V002. This is the latest published version. All versions.

AIFC-072: Company Generation

Status: Draft 0.1 Standard: AI-First Community Standard Abbreviation: AIFC Builds on:

Purpose of this document: To define Company Generation as the process through which AI and people jointly design, validate, create, and prepare a new company or community as an AIFC-compliant system. The document describes company generation from opportunity through purpose, values, operating model, source of truth, Human Cockpit Layer, AI agents, governance, security, fallback, and launch readiness. It also defines boundaries that prevent a generated company from becoming a ghost AI company without human responsibility, values, and real community ownership.

1. Purpose of this document

This document defines Company Generation.

AI can help not only improve an existing company. It can also help design a new company or community as a system.

AI may help identify an opportunity, analyze a market gap, propose purpose, formulate values, describe the customer, propose a business model, create a product or service concept, design an operating model, create workflows, propose roles, prepare a source of truth, create a website, prepare a customer interface, propose AI agents, set governance, prepare a security model, define fallback, and create a launch package.

Company Generation does not mean that AI founds a company by itself.

It means that AI may accelerate the creation of a company as a system, provided that purpose, values, responsibility, decisions, and ownership remain with people or a responsible community.

2. Core principle

The core principle of this document is:

AI may help generate a company system. Humans or a responsible community must own its purpose, values and accountability.

AIFC states:

Generate the system.
Do not generate away responsibility.

Company Generation is acceptable only when a responsible human or community owner exists.

3. Definition

Company Generation is a governed process for designing and preparing a new company or community as an operable system with the help of AI, people, and data sources.

Company Generation may create:

Minimum requirement

Company Generation must have a responsible human or community owner before the generated company begins external operation or accepts customers.

4. Why Company Generation matters

Historically, companies often formed through idea, name, website, first customers, process chaos, tool chaos, documentation debt, and late governance.

AI-first Company Generation may reverse that order:

market signal
-> purpose and values
-> operating model
-> source of truth
-> governance and security
-> human and AI capability
-> customer interface
-> launch

This is a fundamental change. A company can be designed as a system before it grows through chaos.

Minimum requirement

Company Generation must begin with purpose, values, and responsibility, not only with a product, website, or automation.

5. Company Generation is not company ownership

AI may generate proposals, structures, alternatives, workflows, texts, models, risks, simulations, and experiments.

AI cannot own purpose, values, responsibility, legal entity, moral decisions, commitments to customers, commitments to employees, or impact on a community.

Minimum requirement

Company Generation must clearly distinguish AI-generated proposals from human or community-owned decisions.

6. Company Generation vs Company as Product

Company as Product describes a productized operating model that can be deployed or licensed.

Company Generation describes the process of creating a new company or community as a system.

Company as Product:
reusable operating model

Company Generation:
process of creating a new company system

Company Generation may use a Company as Product package. It may also create a new operating model from scratch.

Minimum requirement

Company Generation must clearly distinguish whether it creates a new company from scratch, adapts an existing productized model, or combines both.

7. Reverse company founding

AIFC allows the concept of reverse company founding.

Traditional founding often begins with an idea, product, website, first sales, improvisation, and later structure.

Reverse company founding begins with an identified need, value framework, operating model, source of truth, governance, security, AI capacity model, Human Capability Reserve, and only then public launch.

Minimum requirement

Reverse company founding must verify that the generated system has real purpose and a responsible owner, not only a market opportunity.

8. Opportunity discovery

Company Generation may begin with opportunity discovery.

AI may analyze public data, market gaps, customer frustrations, search trends, public discussions, support patterns, regulation, technological change, local needs, demographic change, and new AI possibilities.

Opportunity discovery must not be only about profit. It must assess whom the service helps, what value it creates, what risks it brings, who bears costs, whether the purpose is ethical, and whether ghost AI company risk appears.

Minimum requirement

Opportunity discovery must assess value, risk, and affected communities, not only market potential.

9. Market gap hypothesis

A market gap hypothesis describes the assumption that an unmet need exists.

It should include target community or customer, problem, current alternatives, why they are insufficient, proposed value, willingness to pay or adoption reason, risks, ethical questions, AI role, and validation.

AI may propose the market gap. People or the community must validate it.

Minimum requirement

A market gap hypothesis must be marked as a hypothesis until it is validated.

10. Purpose generation

AI may help propose a purpose statement.

Purpose must not be only a marketing phrase. It must answer why the company exists, whom it serves, what change it wants to bring, what it rejects, what it must not sacrifice, and how it will know that it makes sense.

Purpose must be approved by the responsible human or community owner.

Minimum requirement

Purpose generated by AI must be reviewed, adapted, and owned by a person or community.

11. Values generation

AI may help propose a values model.

The values model must include values, practical boundaries, non-negotiables, conflicts, decision principles, AI boundaries, customer promises, employee promises, vendor boundaries, and social or environmental boundaries where relevant.

Values are not decoration. They must be usable for decision-making.

Minimum requirement

Generated values must be translated into decision boundaries and approved by a responsible owner.

12. Business model generation

AI may help propose a business model.

The business model must address customer, value, revenue, cost, delivery model, support model, AI cost, human cost, margin, scalability, dependency, risk, compliance, and customer trust.

An AI-generated business model must be validated against reality.

Minimum requirement

A business model generated by AI must include explicit assumptions and a validation plan.

13. Customer model generation

Company Generation must create a customer model.

The customer model describes who the customer is, what problem the customer has, what values and concerns the customer has, what the customer expects from people and AI, how trust is handled, how support works, how feedback is received, and how data is protected.

AI may infer a customer model, but real customer discovery must validate reality.

Minimum requirement

The customer model must distinguish AI inference from validated customer knowledge.

14. Product or service generation

AI may propose a product or service concept.

It must include core offering, value proposition, key features, service boundaries, customer journey, support model, AI involvement, data handling, quality criteria, pricing hypothesis, delivery constraints, and failure modes.

Minimum requirement

The generated product or service concept must have clear boundaries, quality criteria, and support implications.

15. Operating model generation

The operating model is the core of Company Generation.

It must describe roles, workflows, decision rights, support, maintenance, learning, governance, tools, source of truth, AI agents, human skills, AI skills, escalation, security, feedback loops, and metrics.

AI may propose the operating model. People must assess whether it is operable.

Minimum requirement

The generated operating model must include delivery, support, maintenance, governance, security, learning, and fallback.

16. Source of truth generation

Company Generation must create the initial source of truth.

It may include:

/purpose
/values
/strategy
/current-state
/desired-state
/path
/decisions
/workflows
/skills
/ai-governance
/security
/interfaces
/feedback
/risks
/support
/maintenance

The source of truth must be human-readable, agent-actionable, software-verifiable where relevant, versioned, classified, owned, and reviewable.

Minimum requirement

The generated company must have an initial source of truth before significant AI-agentic operation.

17. Human Cockpit Layer generation

AI may help design the Human Cockpit Layer.

The cockpit may show purpose, values, current state, desired state, path, launch readiness, open assumptions, risks, needed decisions, AI agents, budget, feedback, support signals, maintenance tasks, and security alerts.

The cockpit must not be only attractive UI. It must connect to the source of truth.

Minimum requirement

The generated Human Cockpit Layer must be connected to source of truth and governance.

18. AI agent generation

Company Generation may propose AI agents such as market research agent, customer feedback agent, support agent, content agent, sales assistant, operations agent, knowledge maintenance agent, security assistant, or governance assistant.

Each agent must have role, owner, purpose, scope, permissions, forbidden actions, AI skill, cost guardrails, audit, fallback, and offboarding.

Minimum requirement

Generated AI agents must not be activated without permissions, owner, AI-NDA Boundary, and audit.

19. Human role generation

Company Generation must propose human roles. It must not assume that AI will do everything.

Human roles may include company owner, values owner, customer owner, product owner, operations owner, knowledge owner, AI governance owner, security owner, support owner, finance owner, legal or compliance owner, and human review roles.

Minimum requirement

The generated company must define human roles required for accountability, review, support, governance, and fallback.

20. Skill generation

Company Generation must create human skills and AI skills.

Human skills explain how to operate the service, review AI, perform fallback, handle support, work with source of truth, and make decisions.

AI skills explain how agents generate outputs, follow format, avoid forbidden actions, escalate, mark uncertainty, and write drafts.

Minimum requirement

The generated company must include critical human skills and AI skills as paired capabilities.

21. Governance generation

The generated company must have a governance model.

Governance must state who decides, what AI may and may not do, who approves changes, how values conflicts are resolved, how agents are changed, how the operating model changes, how incidents are handled, how audit is kept, and when retrospectives happen.

Minimum requirement

The generated company must include decision governance, AI governance, security governance, and change governance.

22. Security generation

Company Generation must create a security model.

It must include data classification, access control, agent permissions, auditability, AI-NDA Boundary, Operational DNA protection, secrets management, export control, public release review, incident response, backup and recovery, and vendor boundary.

Minimum requirement

The generated company must not launch with non-public knowledge or AI agents without a security model.

AI may help identify legal and compliance questions such as business registration, contracts, privacy, data protection, consumer law, employment, IP, licensing, AI disclosure, taxation, regulated domains, and vendor terms.

AI is not a lawyer with responsibility. Generated legal or compliance content must be reviewed by a responsible person or expert according to risk.

Minimum requirement

The generated company must identify legal and compliance assumptions and required expert reviews before launch.

24. Financial model generation

Company Generation may create a financial model.

It must include revenue hypothesis, cost model, AI cost, human cost, tool cost, vendor cost, customer acquisition cost, support cost, maintenance cost, security cost, margin, runway, sensitivity, and budget guardrails.

Minimum requirement

The generated financial model must include AI cost, human review cost, support cost, and maintenance cost.

25. AI capacity model generation

The generated company must have an AI capacity model.

It states where AI helps, how much AI capacity is needed, what budget exists, what operating mode applies, what cost thresholds exist, what happens when capacity is exhausted, what runs without AI, and how AI waste is monitored.

Minimum requirement

The generated company must define AI capacity, AI budget, operating modes, and reduced-AI behavior.

26. Fallback generation

Company Generation must create a fallback model.

Fallback states what happens without AI, without a specific vendor, without tokens, during a security incident, after agent revocation, during customer escalation, and under minimum viable operation.

Minimum requirement

The generated company must define fallback for critical workflows before launch.

27. Launch readiness

A generated company must not launch only because it looks finished.

Launch readiness must assess whether purpose is approved, values are approved, owner is assigned, operating model is reviewed, source of truth is initialized, security model is ready, AI agents have permissions, customer interface is ready, support is ready, legal assumptions are reviewed, financial assumptions are stated, fallback is ready, and ghost AI company risk has been assessed.

Minimum requirement

The generated company must pass launch readiness review before external operation.

28. Human/community ownership

Company Generation ends when ownership is accepted by a person or community.

Ownership means: we understand the purpose, accept the values, know the risks, know what AI does and must not do, can stop agents, can perform fallback, and accept responsibility toward customers, employees, and affected communities.

Minimum requirement

A generated company cannot be considered AIFC-ready until human or community ownership is explicit and accepted.

29. Company Generation and ghost AI company risk

Company Generation has high ghost AI company risk.

Risk appears when AI creates a brand, website, offer, chatbot, support, automated workflows, generated content, and customer interface, while real owner, responsibility, values, governance, fallback, support, human review, source of truth, and transparency are missing.

Minimum requirement

Company Generation must include ghost AI company risk assessment.

30. Company Generation and represented communities

A new company may affect communities that are not directly at the table, such as children, a local community, future customers, vulnerable groups, the environment, future generations, or data subjects.

AI may help identify affected and represented communities. People must decide how their interests are considered.

Minimum requirement

The generated company must identify affected communities and represented communities where relevant.

31. Company Generation lifecycle

AIFC recommends this lifecycle:

opportunity_discovered
-> hypothesis_created
-> purpose_defined
-> values_defined
-> model_generated
-> model_reviewed
-> security_defined
-> human_ownership_assigned
-> launch_readiness_review
-> launched
-> operated
-> retrospected
-> updated
-> retired

Minimum requirement

The Company Generation project must have lifecycle status and owner.

32. Generated company package

A generated company package may contain:

opportunity hypothesis
purpose statement
values and boundaries
business model
customer model
product/service model
operating model
source of truth structure
Human Cockpit concept
workflow library
role model
human skills
AI skills
AI agents
agent permissions
governance model
security model
data classification policy
access control policy
audit policy
support model
maintenance model
feedback loop
AI capacity model
fallback model
financial model
legal/compliance assumptions
launch readiness checklist
ghost AI company risk assessment

Minimum requirement

The generated company package must include purpose, values, owner, operating model, governance, security, human capability, and fallback.

33. AI role in Company Generation

AI can be very useful in Company Generation.

It may scan opportunities, generate alternatives, create drafts, map risks, prepare models, simulate scenarios, propose agents, propose workflows, prepare documentation, test consistency, and identify gaps.

AI must remain in this role:

generator of proposals
accelerator of design
assistant of validation
not owner of purpose
not owner of values
not owner of responsibility

Minimum requirement

AI role in Company Generation must be documented as proposal generation and decision support, not ownership.

34. Anti-patterns

AIFC rejects the following anti-patterns.

34.1 AI-generated company without owner

AI creates a company, but accountability is unclear.

34.2 Market gap without values

The company is created only from an opportunity, without a value framework.

34.3 Website before operating model

A brand and website appear before governance, support, fallback, and source of truth exist.

34.4 Agents before permissions

AI agents are activated before they have owner, scope, permissions, and audit.

34.5 Business model without support model

The company counts revenue but ignores customer support, complaints, and operational reality.

34.6 AI cost ignored

The model looks profitable but ignores AI cost, human review cost, and maintenance cost.

34.7 No AI-off mode

The company functions only when AI is available.

34.8 Ghost AI company

A company facade is created without a truly responsible community.

AI generates legal-looking content without marking assumptions and review needs.

34.10 Operational DNA unprotected

The generated source of truth contains critical know-how without a security model.

34.11 Customer model only inferred

AI infers the customer, but no one validates it.

34.12 Human roles missing

The company is designed as automations without people responsible for decisions, support, and fallback.

35. Minimal requirements

AIFC Company Generation must at minimum:

  1. Have a responsible human or community owner before external operation.
  2. Begin with purpose, values, and responsibility, not only with product or website.
  3. Distinguish AI-generated proposals from human or community-owned decisions.
  4. Distinguish new company creation, Company as Product adaptation, or a combination.
  5. Verify real purpose and ownership in reverse company founding.
  6. Assess value, risk, and affected communities in opportunity discovery.
  7. Mark the market gap hypothesis as a hypothesis until validation.
  8. Ensure AI-generated purpose is reviewed and owned by a person or community.
  9. Translate generated values into decision boundaries and approve them.
  10. Include assumptions and a validation plan in the business model.
  11. Distinguish AI inference from validated customer knowledge.
  12. Define product or service boundaries, quality criteria, and support implications.
  13. Include delivery, support, maintenance, governance, security, learning, and fallback in the operating model.
  14. Create an initial source of truth before significant agentic operation.
  15. Connect the Human Cockpit Layer to source of truth and governance.
  16. Avoid activating generated AI agents without permissions, owner, AI-NDA Boundary, and audit.
  17. Define human roles for accountability, review, support, governance, and fallback.
  18. Include critical human skills and AI skills as paired capabilities.
  19. Include decision governance, AI governance, security governance, and change governance.
  20. Avoid launching non-public knowledge or AI agents without a security model.
  21. Identify legal and compliance assumptions and required reviews.
  22. Include AI cost, human review cost, support cost, and maintenance cost in the financial model.
  23. Define AI capacity, AI budget, operating modes, and reduced-AI behavior.
  24. Define fallback for critical workflows.
  25. Pass launch readiness review before external operation.
  26. Make human or community ownership explicit and accepted.
  27. Include ghost AI company risk assessment.
  28. Identify affected and represented communities where relevant.
  29. Have lifecycle status and owner.
  30. Include purpose, values, owner, operating model, governance, security, human capability, and fallback in the generated company package.
  31. Document AI role as proposal generation and decision support, not ownership.

36. Summary

Company Generation is the process in which AI helps design a new company or community as a system.

It can change how companies come into existence.

Instead of chaotic growth from idea to tools and later documentation, a company can begin with purpose, values, operating model, source of truth, Human Cockpit Layer, AI governance, security, human skills, AI skills, support, maintenance, feedback, fallback, and a responsible owner.

AIFC states:

AI may generate the company system.
Humans must own the company purpose.
Communities must own the responsibility.

Company Generation is an opportunity to create companies faster, better, and with more structure.

Without governance, it can become a factory for ghost AI companies.

With AIFC, it can become a way to create AI-first, human-managed, purpose-driven companies from the beginning.

Company Generation turns AI-assisted design into responsible company creation.