Insight: Perspectives on Managing AI Ethics in the Digital Age

Winners Consulting Services Co. Ltd. (積穗科研股份有限公司), Taiwan's expert in AI Governance, identifies a critical warning for enterprise leaders in Asia: when algorithmic systems accumulate decision-making power without robust ethical oversight, the result is not just compliance failure — it is a structural threat to democratic accountability and long-term business trust. A landmark 2025 paper published in OpenAlex AI Governance, authored by leading scholar Lorenzo Ricciardi Celsi (h-index: 14, 389 cumulative citations) and Albert Y. Zomaya of the University of Sydney, provides a comprehensive global analysis of AI governance across six major jurisdictions and delivers actionable strategic guidance for Chief Data Officers and AI managers. Taiwan's business leaders cannot afford to ignore its findings.

About the Authors and This Research

Lorenzo Ricciardi Celsi is a prominent European scholar specializing in AI ethics, data governance, and digital transformation. With an h-index of 14 and 389 cumulative academic citations, his work consistently bridges the gap between abstract ethical principles and operational governance frameworks. His co-author, Albert Y. Zomaya, serves as Director of the Centre for Distributed and High Performance Computing at the University of Sydney — one of the world's foremost authorities in computational systems and large-scale distributed AI infrastructure.

The collaboration between these two scholars produces a paper that is rare in its scope: it combines humanistic ethical analysis with technical systems thinking, and it translates academic insight into practical strategic guidance for enterprise decision-makers. Since its publication in 2025, the paper has already attracted 17 citations within the AI governance research community — a strong early indicator of its relevance and impact. For Taiwan's enterprise executives navigating the convergence of ISO 42001, the EU AI Act, and Taiwan's own emerging AI Basic Law, this paper offers an indispensable analytical foundation.

From Algocracy to Algor-Ethics: The Core Conceptual Framework That Every AI Leader Needs to Understand

The paper's central intellectual contribution lies in two opposing yet interdependent concepts: "Algocracy" — the concentration of power in those who control data and algorithms — and "Algor-ethics" — a framework for embedding ethical considerations throughout the entire AI lifecycle. These concepts reframe AI governance not as a technical compliance exercise, but as a fundamental question of institutional design and democratic accountability.

Core Finding One: Algocracy Is the Most Underestimated Institutional Risk in Enterprise AI

The paper defines Algocracy as a structural condition in which the controllers of algorithms and data effectively govern decision-making processes — often without transparency, accountability, or meaningful human oversight. For enterprises, this manifests as algorithmic bias in hiring, credit scoring, or medical triage; as opacity in automated decisions that affect customers or employees; and as accountability gaps that expose organizations to regulatory action under the EU AI Act and Taiwan's AI Basic Law. The paper specifically highlights high-impact domains — healthcare, Insurtech, environmental sustainability, and space exploration — as sectors where the risks of unmanaged Algocracy are most acute. The practical implication for Taiwan's enterprises is unambiguous: AI systems must be governed with human oversight mechanisms built in by design, not added as an afterthought. This is precisely what ISO 42001 mandates.

Core Finding Two: Global Regulatory Divergence Creates Real Strategic Risk for Taiwan's Export-Oriented Businesses

One of the paper's most valuable contributions is its comparative analysis of AI governance across six major jurisdictions. The EU AI Act adopts a risk-based regulatory architecture — classifying AI applications into four tiers: unacceptable risk, high risk, limited risk, and minimal risk — with high-risk applications facing stringent requirements for transparency, human oversight, and auditability. Japan and Canada similarly prioritize fundamental rights in their AI governance frameworks. The United States leans toward innovation-enabling approaches through executive directives and sector-specific oversight. China integrates AI governance with state ideological imperatives, enforcing compliance with socialist core values. Brazil, while committed to fairness and democratic oversight, has yet to develop institutional depth comparable to the more mature frameworks.

For Taiwan's export-oriented manufacturers, technology companies, and financial institutions with European market exposure, the signal is clear: EU AI Act compliance is not a future consideration — it entered into force in 2024 and is now in its phased implementation period. ISO 42001 certification provides the most internationally recognized demonstration of AI governance capability and will serve as a key differentiator in EU supply chain qualification processes.

Core Finding Three: Transdisciplinary Governance Is the Prerequisite for Sustainable AI Value Creation

The paper argues that effective AI governance cannot be achieved through a single disciplinary lens. Legal, ethical, and technical perspectives must be integrated into a coherent, transdisciplinary governance architecture. For enterprises, this has direct organizational implications: AI governance cannot be siloed within IT departments. It requires structured collaboration between legal, compliance, business, and technology functions — a governance architecture that the paper identifies as essential for AI to genuinely serve long-term value creation rather than short-term automation gains.

What This Research Means for Taiwan's AI Governance Practice

Taiwan's enterprises are at a regulatory inflection point. The EU AI Act entered into force in 2024, with its phased implementation timeline placing increasing compliance obligations on organizations operating in or supplying to the European market. Taiwan's AI Basic Law (人工智慧基本法) is advancing through the Legislative Yuan, establishing foundational principles of human-centricity, transparency, and accountability that directly parallel the ethical framework this paper advocates. And ISO/IEC 42001:2023 — the world's first international standard for AI management systems — provides the practical implementation architecture that connects ethical principles to operational governance.

The paper's findings create three urgent action priorities for Taiwan's enterprise AI leaders. First, conduct an immediate AI risk inventory: map all current AI applications against the EU AI Act's four-tier risk classification to identify which systems face high-risk compliance requirements. Second, initiate ISO 42001 gap analysis: assess the distance between current AI governance practices and the requirements of ISO/IEC 42001:2023, prioritizing transparency mechanisms, human oversight protocols, and accountability frameworks. Third, build cross-functional AI governance structures: establish a formal AI Governance Committee that integrates legal, compliance, technology, and business leadership — the organizational foundation that both this paper and ISO 42001 identify as essential for sustainable AI governance.

How Winners Consulting Services Helps Taiwan Enterprises Build World-Class AI Governance

Winners Consulting Services Co. Ltd. (積穗科研股份有限公司) helps Taiwan enterprises build AI management systems aligned with ISO 42001, EU AI Act requirements, and Taiwan's AI Basic Law. Our approach is not compliance documentation — it is the establishment of living, operational governance mechanisms that create real organizational accountability and international credibility.

1. AI Ethics Risk Inventory and Risk Classification: Using the EU AI Act's four-tier risk framework and ISO 42001 requirements as the assessment foundation, we systematically map your organization's AI applications, identify algorithmic bias risks, transparency gaps, and human oversight deficiencies, and build an enterprise-specific AI Risk Register that serves as the operational backbone of your governance system.
2. ISO 42001 AI Management System Implementation: We design and implement an AI governance architecture tailored to your industry and organizational scale, covering policy development, role and responsibility allocation, lifecycle monitoring mechanisms, human oversight protocols, and continuous improvement processes. Our target: foundational system establishment within 90 days, with certification readiness achieved within 120 days for most mid-sized enterprises.
3. Transdisciplinary AI Governance Committee Design: Directly addressing the paper's core finding on transdisciplinary governance, we help enterprises establish a structured AI Governance Committee that integrates legal, compliance, technology, and business leadership. This ensures AI governance is an institutional commitment embedded in organizational structure — not an IT department side project — and creates the accountability architecture required by both EU AI Act and Taiwan's AI Basic Law.

Frequently Asked Questions

What is "Algocracy" and why should Taiwan's enterprise executives care about it?

Algocracy describes a structural condition in which those who control data and algorithms effectively govern decision-making — often without transparency or meaningful accountability. For Taiwan's enterprises, the practical risk is this: if AI systems operate without human oversight mechanisms, explainability requirements, or audit trails, the organization cannot trace, justify, or remedy AI-driven decisions when they go wrong. This creates regulatory exposure under the EU AI Act and Taiwan's AI Basic Law, and reputational risk with customers and partners. ISO 42001 addresses Algocracy directly by requiring enterprises to establish human oversight protocols, accountability frameworks, and transparency mechanisms as core elements of their AI management system. Winners Consulting recommends starting with high-impact applications — HR recruitment AI, credit scoring, medical support tools — where Algocracy risk is highest.

Does Taiwan's export business need to comply with the EU AI Act, and when?

Yes — if your AI systems, products, or services are used in the EU market, or if you supply AI-enabled components to EU-based companies, EU AI Act obligations apply to you. The Act entered into force in August 2024, with a phased implementation schedule. Prohibited AI practices were prohibited from February 2025; high-risk AI system requirements apply from August 2026 for most categories. Taiwan's technology exporters, OEM/ODM manufacturers incorporating AI features, financial institutions serving EU clients, and logistics companies using AI in EU operations all face potential compliance obligations. The critical first step is an AI application risk classification exercise — conducted against the EU AI Act's four-tier risk framework — to determine which systems are subject to which requirements. Winners Consulting provides EU AI Act gap analysis services designed specifically for Taiwan enterprises.

What is ISO 42001, and how does it relate to EU AI Act compliance and Taiwan's AI Basic Law?

ISO/IEC 42001:2023 is the world's first international standard for AI management systems. It provides an auditable, certifiable framework for embedding ethical and governance principles throughout the AI lifecycle — covering risk management, transparency, human oversight, accountability, and continuous improvement. ISO 42001 certification serves multiple compliance purposes simultaneously: it demonstrates EU AI Act governance capability to European regulators and supply chain partners; it aligns with the "human-centric," "transparent," and "accountable