Part II — Classifying EUREKA Across the AI Value Chain and Layers of AI

EUREKA: Ethical AI Framework

Most AI frameworks begin with values. Few explain where values should operate inside the AI ecosystem. This creates a recurring governance gap: Organizations publish ethical principles yet struggle to operationalize them across infrastructure, models, applications, and decision systems.

The EUREKA Framework addresses this gap. Its latest evolution is not simply an ethical declaration. It becomes a layered governance architecture that can be embedded throughout the AI value chain. This section therefore answers a strategic question: At what layer of AI does EUREKA create governance value?

The answer: Every layer.

Because AI risk does not emerge only at deployment. Risk accumulates throughout the chain.

Understanding Where AI Creates—and Concentrates—Power

The AI ecosystem can be understood as five interdependent layers:

Hardware sits at the bottom of the AI value chain.

Without chips, data centers, energy systems, connectivity, and devices—AI does not exist.

Yet governance discussions often ignore infrastructure. This is a mistake. Infrastructure determines who can innovate; who controls computation; who accesses AI; who absorbs environmental costs. Hardware governance therefore becomes an ethical question.

EUREKA Classification:

Primary Principles:
E — Ethical
E — Equitable
A — Accountable

Governance Questions

Ethical:

• Are infrastructure investments socially beneficial?

Equitable:

• Who gains access?

Accountable:

• What environmental and social impacts are measured?

Global AI governance increasingly recognizes sustainability, access, and infrastructure readiness as prerequisites for inclusive digital transformation.

Layer 2 — Cloud Platforms

Responsibility in AI Enablement

Cloud platforms have become the operating layer of modern AI.

They provide:

• compute
• storage
• orchestration
• APIs
• deployment environments

Cloud concentration raises governance questions around:

• dependency
• resilience
• interoperability
• digital sovereignty
• cybersecurity

EUREKA treats cloud not as neutral infrastructure.

Cloud determines governance capacity.

EUREKA Classification:

Primary Principles:
R — Responsible
A — Accountable
E — Equitable

Governance Controls

Responsible:

• deployment safeguards

Accountable:

• audit logs

Equitable:

• accessible infrastructure

Cloud governance becomes public-interest governance.

Layer 3 — Data and AI Models

The Core Governance Layer of EUREKA

This is where AI becomes intelligent.

Data and models determine:

• learning
• prediction
• generation
• recommendation
• autonomy

This layer produces the greatest concentration of governance risk.

Typical failures include:

• biased datasets
• model opacity
• hallucinations
• weak explainability
• unsafe optimization

This is where EUREKA becomes most operational.

EUREKA Classification:

Primary Principles:
U — Understandable
K — Knowledge-Driven
R — Responsible
A — Accountable

Governance Controls

Understandable:

• explainability

Knowledge-Driven:

• evidence quality

Responsible:

• lifecycle management

Accountable:

• documentation

AI governance literature increasingly emphasizes governance across the entire AI lifecycle rather than regulating outputs alone.

Where AI Meets Humanity

Applications are where society experiences AI.

Examples include:

• education systems
• health platforms
• public services
• hiring tools
• legal systems
• financial technologies

Applications convert technical decisions into social outcomes.

This is where ethical failure becomes visible.

EUREKA Classification:

Primary Principles:
E — Ethical
E — Equitable
U — Understandable

Governance Controls

Ethical:

• human dignity

Equitable:

• inclusion

Understandable:

• explainable experiences

Applications determine public trust.

Trust determines adoption.

Adoption determines national competitiveness.

Layer 5 — Governance and Society

The Meta-Layer of Accountability

Governance is not above AI.

Governance surrounds AI.

This layer includes:

• law
• standards
• institutions
• ethics
• assurance
• international coordination

This is where EUREKA becomes a societal architecture.

EUREKA Classification:

Primary Principles:
A — Accountable
R — Responsible
K — Knowledge-Driven

It aligns with—and extends—major international approaches.

1. UN Governance Layer → EUREKA as Human-Centered Coordination

Recent UN initiatives increasingly position AI as a global public-interest issue.

Relevant governance references include:

• UN High-Level Advisory Body on AI (2024) — inclusive global coordination
• UN System White Paper on AI Governance (2024) — policy coherence
• UNESCO Recommendation on the Ethics of AI (2021) — ethics and capacity-building
• Global Digital Compact (2024) — human-centered digital governance

EUREKA Contribution:

Moves from international norms → personal and institutional conduct.

This makes EUREKA operational at practitioner level.

2. Rights-Based Governance → EUREKA as Applied Human Rights

Rights-based models emphasize:

• dignity
• equality
• transparency
• democratic oversight

Examples include:

• Council of Europe Convention on AI and Human Rights
• OECD AI Principles
• Global Partnership on AI

EUREKA Contribution:

Converts rights language into actionable professional decisions.

3. Regional Governance → EUREKA as Contextual Governance

Regional approaches increasingly recognize that AI governance cannot be copied universally.

Examples include:

• ASEAN Guide on AI Governance and Ethics
• Ibero-American AI Charter
• African Union Continental AI Strategy

These emphasize:

• capacity
• interoperability
• regional readiness
• inclusion

EUREKA Contribution:

Creates a Filipino and Asia-Pacific interpretation of ethical AI leadership.

4. Risk-Based Regulation → EUREKA as Governance-by-Layer

Risk-based approaches classify AI according to impact.

Examples include:

• EU Digital Services Act
• G7 Hiroshima Process Code of Conduct

These emphasize:

• transparency
• safety
• accountability
• mitigation

EUREKA Contribution:

Adds a missing dimension:

Not only:
“What is the risk?”

But:

“What values should govern each layer?”

The final insight of Part II is this:

EUREKA should not be treated as another checklist.

It should be positioned as:

A Cross-Layer Human Governance Architecture

Its function is to connect:

Infrastructure → Systems → Decisions → Society

Through six commitments:

Ethical
Understandable
Responsible
Equitable
Knowledge-Driven
Accountable

If Part I asked:

What values should govern AI?

Part II answers:

Where should those values operate?

The answer is:

Across the entire AI value chain.

EUREKA therefore evolves from a code of conduct into a governance architecture—one capable of guiding professionals, institutions, and nations toward trustworthy, inclusive, and human-centered AI.

Pages: 1 2 3 4