Human-Centered Governance Frameworks for Artificial Intelligence

A flagship research collection on designing AI governance around human judgment, accountability, and auditable oversight.

Abstract—This flagship report compiles four governance-oriented whitepapers grounded in Atmakosh-aligned principles. The collection advances a human-centered approach to artificial intelligence that prioritizes institutional legitimacy, plural reasoning, accountable decision support, and non-ideological ethics. The report contains: (1) Plural Intelligence, (2) Governance-First AI, (3) Decision Systems for Boards, and (4) Ethics Without Ideology.

Index Terms—governance, plural intelligence, board decision-making, ethics, institutional legitimacy, human-centered AI

I. PLURAL INTELLIGENCE

Abstract

Plural Intelligence reframes artificial intelligence for envi-ronments where legitimate disagreement is permanent. Rather than treating variance as noise, Plural Intelligence treats dif-ference as a resource for institutional resilience. Grounded in Atmakosh-aligned principles of coherence, humility, and service to human judgment, this paper proposes a governance-oriented approach in which AI supports understanding across perspectives without collapsing them into a single doctrine, metric, or narrative. The result is an intelligence posture suitable for boards, policy bodies, and public institutions whose legitimacy depends on navigating plural values.

Index Terms

plural intelligence, governance, institutional legitimacy, de-liberation, human agency, non-technical AI ethics

A. Introduction

Modern institutions operate in plural conditions: multiple stakeholder groups, competing value commitments, uncertain futures, and contested interpretations of events. In such en-vironments, the most dangerous failure mode is not mere inaccuracy; it is false certainty presented as neutrality. When an intelligence system produces a single authoritative output, it can convert a reasonable plurality of views into an apparent unanimity. This dynamic erodes trust, marginalizes minority perspectives, and encourages brittle decisions. Plural Intel-ligence addresses this gap by shifting the role of AI from answer production to perspective stewardship. It emphasizes that the purpose of intelligence in governance settings is not to eliminate disagreement but to make disagreement legible, bounded, and constructive. Within Atmakosh-aligned framing, the system’s highest virtue is not dominance but coherence: the capacity to hold differences in a relationship that remains intelligible to human decision-makers.

B. From Singular Optimization to Plural Coherence

Singular optimization is attractive because it promises clarity. Yet in governance settings, clarity purchased by exclusion is a liability. A system that optimizes for one metric will tend to subordinate other values: fairness, autonomy, dignity, long-term stability, cultural context, or institutional legitimacy. The outcome may look efficient, but it becomes fragile when conditions change or when neglected stakeholders contest the decision. Plural coherence replaces the question “What is the best answer?” with “What are the meaningful answers, for whom, under what assumptions, and with what trade-offs?” The system surfaces the space of reasonable interpretations and clarifies how each interpretation depends on values and context. This is not indecision; it is disciplined awareness.

C. Atmakosh-Aligned Principles for Plural Intelligence

Plural Intelligence can be grounded in a small set of non-technical principles. Coherence before conclusion: The system prioritizes internal consistency and explicit trade-offs, resisting premature closure. Humility and boundedness: Outputs are framed as perspectives and scenarios, not declarations. The system acknowledges uncertainty. Service to human insti-tutions: The system supports deliberation while preserving accountability in human governance bodies. Respect for plu-rality: Divergent values are treated as legitimate inputs rather than errors to be corrected. Contestability: The system invites challenge, alternative framing, and structured disagreement.

D. Institutional Use Cases

Plural Intelligence is most valuable when legitimacy de-pends on recognizing multiple viewpoints. Board and exec-utive deliberation: When strategic choices involve trade-offs (growth vs. risk, innovation vs. trust), Plural Intelligence helps leadership articulate multiple plausible readings of the same situation and prevents a single narrative from hardening too early. Public consultation and participatory policy: Stake-holder submissions often contain conflicting priorities. A plu-ral system can help summarize disagreements faithfully with-out flattening them into a misleading average. Diplomacy and cross-cultural engagement: Plural Intelligence can preserve cultural nuance by presenting multiple legitimate interpreta-tions rather than assuming one universal frame. Ethical review and impact assessments: Ethics committees rarely seek a single correct answer; they seek a defensible, transparent balance. Plural Intelligence surfaces options, consequences, and moral tensions in a structured way. Plural Intelligence can

be understood through the lens of institutional “sensemaking.” Sensemaking is the disciplined practice by which a group interprets ambiguous signals, negotiates meaning, and forms a shared basis for action. In high-stakes contexts, sensemaking fails when a single storyline becomes prematurely dominant. Plural Intelligence protects sensemaking by keeping alternative frames visible long enough for scrutiny and comparison. An Atmakosh-aligned approach treats plurality as a stability mechanism. When multiple perspectives are documented, an institution can revisit its reasoning after outcomes unfold. This creates an audit trail of judgment: not merely what was decided, but what was considered, what was feared, and what was valued. Such traceability is crucial for legitimacy, espe-cially when decisions are contested. A practical way to adopt Plural Intelligence is to define a small set of recurring “per-spective registers” appropriate to the institution. For example, a public agency may require: (i) rights and dignity, (ii) service equity, (iii) fiscal stewardship, (iv) operational feasibility, and (v) public trust. A board may require: (i) strategic resilience, (ii) regulatory exposure, (iii) cultural impact, (iv) stakeholder legitimacy, and (v) long-horizon value creation. The system’s role is to populate each register with a faithful articulation, not to collapse registers into one score. Plural Intelligence also supports healthier disagreement. Institutions often confuse disagreement with dysfunction. In reality, disagreement be-comes dysfunction only when it becomes unmanaged: hidden, personal, or unstructured. By presenting disagreements as structured trade-offs, the system converts conflict from identity threat into governance material. Finally, Plural Intelligence is compatible with decisiveness. The goal is not to extend deliberation indefinitely, but to ensure that when a decision is made, it is made with awareness of who benefits, who bears cost, and what uncertainty remains. This is how institutions earn the right to act.

E. Governance Requirements

In plural settings, the main risk is that a system becomes an unaccountable narrator. Governance requirements therefore focus on role clarity and institutional control. Role definition: The system must be explicitly framed as an advisory instrument, not a decision-maker. Human authorities retain final responsibility. Boundary conditions: Institutional policy should define where plural outputs are required (e.g., policy deliberation) and where single outputs are appropriate (e.g., procedural compliance). Documentation of assumptions: For every perspective presented, the system should state the value premises and contextual constraints that make it reasonable. Right to challenge: Stakeholders should be able to contest how their viewpoint was represented and request revision.

F. Risk Analysis and Failure Modes

Plural Intelligence introduces its own risks if poorly gov-erned. False balance: Presenting all perspectives as equally credible, even when some are demonstrably harmful or inco-herent. The mitigation is not ideology but standards: reject perspectives that advocate violence, deny basic human dig-nity, or deliberately misrepresent evidence. Decision deferral:

Leaders may hide behind plurality to avoid responsibility. Mitigation requires explicit decision deadlines and account-ability assignment. Narrative capture: Powerful stakeholders may pressure the system to frame outputs to their advantage. Mitigation requires independent oversight and transparency about representation. Overcomplexity: Excessive perspectives can overwhelm decision-makers. Mitigation requires curated plurality: surface the most relevant frames and allow deeper exploration when needed.

G. Evaluation and Readiness Criteria

Plural Intelligence should be evaluated on institutional outcomes rather than technical benchmarks. Faithful repre-sentation: Do stakeholders recognize their position in the system’s articulation? Trade-off clarity: Are value conflicts made explicit rather than hidden? Decision quality: Do boards and policy bodies report fewer surprises and more robust reasoning? Legitimacy signals: Do affected groups perceive decisions as more explainable and contestable? Resilience: When conditions shift, can the institution adapt without losing coherence?

H. Conclusion

Plural Intelligence is not a luxury feature; it is a governance necessity for AI deployed in plural societies. By shifting from singular answers to structured perspectives, institutions gain resilience, legitimacy, and better judgment. Grounded in Atmakosh-aligned principles of humility and coherence, Plural Intelligence offers a path toward AI that strengthens deliberation rather than replacing it.

II. GOVERNANCE-FIRST AI

Abstract

Governance-First AI reverses the prevailing “deploy then regulate” pattern by treating governance as the first design decision and the last operational responsibility. Grounded in Atmakosh-aligned principles of restraint, accountability, and human institutional sovereignty, this paper presents a non-technical governance blueprint for organizations and public bodies adopting AI. The central claim is simple: capability without legitimacy becomes institutional risk, while gover-nance as a foundational layer converts risk into trust and trust into sustainable adoption.

Index Terms

governance-first, accountability, institutional legitimacy, oversight, contestability, risk management

A. Introduction

AI is increasingly used in contexts where errors do not merely degrade performance; they reshape rights, opportu-nities, and public trust. Yet the dominant pattern remains capability-first: systems are adopted for efficiency, then gover-nance is retrofitted after visible harm. This creates predictable failure: diffuse accountability, unclear boundaries, and erosion of legitimacy. Governance-First AI argues that governance is not a compliance layer. It is the operating constitution of AI use. If the institution cannot explain who owns the outcomes, what the system is for, what it must never do, and how disagreements are resolved, the system is not ready for consequential use.

B. The Capability-First Failure Pattern

Three structural dynamics make retrofitted governance un-reliable. Path dependence: Once embedded in workflows, AI becomes difficult to unwind even when harms are discovered. Responsibility drift: When outputs appear authoritative, hu-man decision-makers may defer, creating a gap between formal responsibility and practical behavior. Legitimacy debt: Stake-holders experience decisions as imposed rather than reasoned, reducing cooperation and increasing conflict. Governance-first design prevents these dynamics by establishing the rules of en-gagement before the system participates in decision processes.

C. Atmakosh-Aligned Governance Principles

Governance-first can be framed through a set of princi-ples that avoid technical prescription. Purpose boundedness: Every deployment must have a declared purpose, prohibited uses, and defined limits. Human accountability: A named human authority owns outcomes. “The system decided” is never acceptable. Contestability: Affected parties can question outputs and request review. Proportionality: The stronger the impact, the stronger the governance, review, and escalation requirements. Operational humility: The institution treats AI as fallible and revisable, not final.

D. Governance Architecture for Institutions

A practical governance architecture can be organized into four layers. Mandate layer: Defines why the system exists, what it serves, and what it must not do. Authority layer: Assigns decision ownership, escalation pathways, and review responsibilities. Oversight layer: Establishes independent re-view, periodic audits, and incident response. Transparency layer: Communicates scope, limitations, and contestability to stakeholders. These layers can be implemented through policy, training, and governance committees without disclosing any technical design. Governance-first should be treated as a leadership discipline rather than a documentation exercise. The central question for executives is: “If something goes wrong, will our institution respond with clarity, responsibility, and learning—or with denial and confusion?” Governance-first practices create the conditions for learning under pressure. A governance-first program typically begins with a charter. The charter states the system’s permitted purpose, prohibited uses, escalation routes, and the identity of accountable leaders. Importantly, the charter is written in plain language so that non-specialists can understand what the institution is doing. Transparency is a legitimacy strategy. Second, governance-first requires a culture of non-deference. Many organizations adopt tools but fail to train people to challenge them. Governance-first training includes rituals such as “assumption checks,” “counterargument rounds,” and “pause authority”—the explicit permission to stop a process when uncertainty or harm risk rises. Third, governance-first treats stakeholder communication as part of governance. If affected groups do not understand the system’s role, they cannot meaningfully contest it. Institutions should therefore publish scope statements, provide accessible complaint channels, and demonstrate responsive correction when concerns are raised. Fourth, governance-first strengthens resilience by preparing for incidents. Incident preparation is not pessimism; it is maturity. Institutions should define how they will identify harmful outcomes, who will investigate, how remediation will occur, and how learning will be institution-alized. A credible incident practice deters panic and protects trust.

E. Lifecycle Governance

Governance-first is continuous. Key checkpoints include: Pre-adoption: risk assessment, stakeholder mapping, purpose declaration. Pilot: limited scope, heightened review, docu-mentation of observed failure modes. Scale: formal oversight cadence, incident handling, public communication strategy. Revision/retirement: criteria for pausing, redesigning, or ending use. Lifecycle governance prevents the illusion that a one-time approval can cover long-term evolving risks.

F. Risk and Opportunity Summary

Opportunities: stronger institutional trust; better regulatory readiness; reduced reputational shocks; clearer decision own-ership. Risks: governance theater (symbolic processes); ac-countability diffusion; over-reliance on policy without culture change. Mitigation requires aligning governance documents

with real decision practices, training, and leadership account-ability.

G. Evaluation Metrics for Governance-First

Institutions can evaluate governance-first adoption through observable indicators. Clarity: Can leaders explain the sys-tem’s purpose and prohibited uses in plain language? Owner-ship: Is a responsible authority named for each consequential use? Contestability uptake: Do people actually use challenge pathways, and do those pathways work? Incident maturity: Are failures documented, learned from, and corrected without denial? Legitimacy: Do stakeholders report increased trust and understanding?

H. Conclusion

Governance-First AI is not slower innovation; it is safer acceleration. By placing legitimacy and accountability be-fore capability, institutions avoid predictable crises and build durable trust. Atmakosh-aligned governance empha-sizes humility, bounded purpose, and human institutional sovereignty—ensuring AI strengthens governance rather than quietly replacing it.

III. DECISION SYSTEMS FOR BOARDS

Abstract

Board decisions combine uncertainty, high stakes, and long-term consequences. Traditional AI adoption often emphasizes speed and certainty, but board governance requires delib-eration, accountability, and ethical foresight. Grounded in Atmakosh-aligned principles of service to human judgment and institutional integrity, this paper proposes a non-technical model of decision systems for boards: AI as an insight layer that expands perspective, surfaces trade-offs, and reveals second-and third-order impacts, while directors retain full fiduciary responsibility.

Index Terms

board governance, decision support, fiduciary duty, fore-sight, ethics, institutional risk

A. Introduction

Boards exist to govern the future under uncertainty. They are accountable not only for financial outcomes but also for legitimacy, trust, and stewardship. Yet many AI tools entering boardrooms were designed for operational optimization: they compress complexity into a score, a recommendation, or a predicted outcome. In board contexts, this compression can be dangerous. It may hide assumptions, understate ethical risk, or create a false impression that the “right” decision is the one the system presents. A board decision system should instead function like a disciplined advisor: it clarifies the landscape of choices, makes assumptions visible, and helps directors reason across time horizons and stakeholder impacts. This posture aligns with Atmakosh principles by prioritizing coherence and responsibility over automation.

B. Why Board Decisions Differ

Board decisions have four distinguishing features. Tempo-ral depth: Consequences unfold over years, not weeks. Stake-holder breadth: Decisions affect employees, customers, com-munities, regulators, and the institution’s legitimacy. Moral loading: Many strategic choices contain ethical trade-offs (privacy vs. personalization, efficiency vs. inclusion). Ac-countability: Directors cannot delegate fiduciary duty to a tool. Therefore, board decision support must prioritize inter-pretability, contestability, and trade-off visibility.

C. AI as an Insight Layer, Not a Decision Engine

In an Atmakosh-aligned framing, the system’s role is to improve the quality of human judgment. It does so by: Mapping options: present the viable strategic paths, including “do nothing” and “pause.” Surfacing assumptions: state what must be true for each option to succeed. Projecting conse-quences: outline plausible second-and third-order effects on trust, regulation, culture, and resilience. Revealing trade-offs: identify what is gained and what is sacrificed. Encouraging dissent: present counterarguments and minority viewpoints. The system does not finalize decisions; it makes decision-making more conscious.

D. Decision Landscapes and Scenario Discipline

A board decision is rarely about one variable. It is about a landscape: interacting uncertainties, constraints, and values. A decision system should therefore support scenario discipline. Multiple futures: show several plausible trajectories rather than a single forecast. Stress conditions: highlight what happens when assumptions fail. Reversibility: clarify whether a decision can be undone and at what cost. Governance triggers: propose objective signals that would prompt re-evaluation. Scenario discipline reduces surprise and protects boards from overconfidence.

E. Governance and Control for Board Use

Because board decisions are consequential, governance must be explicit. Authority: outputs are advisory; the chair and board retain responsibility. Minutes and traceability: the board should record which perspectives were considered and why a path was chosen. Conflict-of-interest safeguards: ensure the system’s framing is not captured by any one faction. Confidentiality: protect sensitive deliberation while maintaining accountability. Escalation: define when decisions require external review (e.g., high societal impact). A board decision system should also support governance cadence. Boards operate through agendas, committees, and periodic reviews. The system should therefore align with existing governance rhythms: pre-read preparation, committee deep dives, plenary debate, and post-decision monitoring. The aim is to strengthen, not disrupt, board process. To protect fiduciary integrity, boards can establish “decision hygiene” rules. Examples include: (i) never accept a recommendation without examining assumptions; (ii) always request at least one strong counterargument; (iii) require an explicit articulation of stakeholder impact; and (iv) record uncertainty. These rules transform AI use into disciplined governance rather than convenience. Board decision systems are particularly valuable for long-horizon risk. Traditional reporting favors quarterly indicators, but many failures emerge slowly: cultural degradation, trust erosion, regulatory drift, and capability dependence. A reflective system can continuously surface early signals and invite timely intervention. A key benefit is improved narrative governance. Boards often govern through narratives: the story of what the institution is, what it owes, and where it is going. AI can help test narratives against evidence, alternative interpretations, and ethical implications—without claiming to be the author of institutional meaning. Ultimately, the board’s relationship with AI should mirror good governance: clarity of role, insistence on accountability, openness to challenge, and commitment to learning.

F. Risk and Opportunity Summary

Opportunities: reduced groupthink; improved long-horizon reasoning; clearer ethical visibility; stronger regulatory pre-paredness. Risks: deference to system authority; illusion of precision; misuse as liability shield; information overload. Mit-igation requires training directors to treat outputs as perspec-tives, instituting challenge rituals, and setting strict boundaries on use.

G. Board Readiness Indicators

A board can assess readiness for AI decision support through governance maturity signals. Question quality: Di-rectors ask better questions after using the system. Dissent health: Minority views are invited and documented. Ethical visibility: discussions include societal and human impacts, not only financial metrics. Accountability clarity: no ambiguity about who owns outcomes. Learning loop: decisions are revisited, and lessons are recorded.

H. Conclusion

Decision systems for boards should be designed to strengthen governance, not to accelerate execution. When aligned with Atmakosh principles, AI becomes a disciplined mirror for leadership: it expands perspective, makes trade-offs visible, and supports long-term stewardship. The board remains accountable—yet better equipped to be wise.

IV. ETHICS WITHOUT IDEOLOGY

Abstract

Ethics Without Ideology offers a governance-friendly ap-proach to AI ethics for pluralistic and cross-cultural con-texts. Rather than enforcing a single doctrine, it emphasizes transparency of impact, respect for human agency, and harm reduction as shared ethical ground. Grounded in Atmakosh-aligned principles of coherence, humility, and dignity, this paper proposes an iterative ethical review model suitable for institutions, regulators, and multilateral environments without relying on culturally exclusive assumptions.

Index Terms

AI ethics, harm reduction, human agency, pluralism, gover-nance, institutional legitimacy

A. Introduction

Many ethics frameworks fail not because they are wrong, but because they are incompatible with plural societies. When ethics is framed as ideological enforcement, institutions en-counter resistance, accusations of bias, and a collapse of trust. AI systems deployed across borders intensify this problem: what is assumed morally obvious in one context may be con-tested in another. Ethics Without Ideology seeks a pragmatic ethical core that can travel across contexts. It focuses on what institutions can evaluate: harms, agency, transparency, and dignity. This approach does not abandon moral seriousness; it relocates ethics from doctrinal allegiance to accountable judgment.

B. Why Ideological Ethics Break at Scale

Ideological ethics tends to break in four predictable ways. Exclusion: Ethical conclusions are treated as obvious to insid-ers and illegitimate to outsiders. Rigidity: Frameworks cannot adapt when new harms or contexts emerge. Weaponization: Ethical language becomes a tool for political or institutional dominance. Moral outsourcing: Organizations treat compli-ance with a doctrine as equivalent to ethical responsibility. Ethics Without Ideology avoids these failures by emphasizing process, transparency, and consequence awareness.

C. A Minimal Shared Ethical Core

A non-ideological ethical core can be constructed from widely shareable commitments. Harm reduction: Prefer ac-tions that minimize foreseeable harm to individuals and com-munities. Human agency: Preserve the ability of people to un-derstand, contest, and meaningfully choose. Dignity and non-humiliation: Avoid systems that degrade people, treat them as disposable, or deny their basic standing. Transparency of impact: Make the scope, limitations, and consequences of system use visible. These commitments are not ideological conclusions; they are governance necessities for legitimacy.

D. Atmakosh-Aligned Ethical Posture

Atmakosh-aligned ethics emphasizes inward coherence and humility. Coherence: Ethical reasoning should be consistent with the institution’s declared purpose and values. Humility: Ethical claims must acknowledge uncertainty and the limits of foresight. Service: Ethics exists to protect people and institutions, not to win arguments. Contextuality: The same action can have different ethical meaning in different contexts. This posture discourages moral absolutism while preserving moral seriousness.

E. The Ethics Without Ideology Loop

Ethical governance can be operationalized as an iterative loop. Context: Identify stakeholders, setting, cultural con-straints, and vulnerability. Impact: Map foreseeable bene-fits and harms, including indirect effects. Agency: Evaluate whether people can understand, contest, and opt out. Re-view: Reassess after deployment; incorporate feedback; revise boundaries. The goal is continuous ethical learning rather than one-time certification.

F. Institutional Applications

Regulatory sandboxes: Provide a shared ethical language for pilots without forcing ideological consensus. Public ser-vices: Enable defensible, transparent decisions in welfare, education, and health contexts. Cross-border deployments: Reduce friction by focusing on harms and agency rather than cultural doctrine. Corporate governance: Support board over-sight by translating ethical risk into institutional obligations. Ethics Without Ideology is especially useful where legitimacy depends on inclusiveness. Ethics Without Ideology does not imply ethical minimalism. Rather, it distinguishes between (i) shared governance commitments that enable legitimacy and (ii) deeper moral philosophies that vary across cultures. Insti-tutions can remain ethically serious while avoiding the mistake of treating one tradition as universally binding. A useful opera-tional practice is “ethical translation.” When an ethical concern is raised in ideological language, the institution translates it into the shared core: Where is the harm? Whose agency is constrained? What dignity is compromised? What impacts are hidden? This translation allows diverse stakeholders to delib-erate without requiring agreement on metaphysics or politics. In plural societies, dignity is often the most stable ethical anchor. Systems that humiliate, stereotype, or render people invisible predictably generate resistance and harm. Therefore, dignity checks should be standard: does the system treat persons as ends, not instruments? does it allow meaningful contest? does it avoid framing groups as disposable? Ethical review should also include “distributional attention.” Even when aggregate outcomes improve, harms may concentrate on vulnerable groups. A non-ideological approach makes this visible by requiring explicit description of who bears risk and who benefits. Finally, ethics without ideology is strengthened by institutional courage: the willingness to pause or withdraw from uses that cannot be made legitimate. Ethical governance is not only about mitigation; it is about boundaries.

G. Risk and Opportunity Summary

Opportunities: cross-cultural viability; reduced ideological conflict; stronger stakeholder trust; improved ethical learning loops. Risks: superficial minimalism; avoidance of hard moral questions; inconsistent application across units. Mitigation requires strong review processes, transparent reasoning, and explicit escalation for high-impact cases.

H. Evaluation Criteria

Institutions can evaluate this approach through measurable governance outcomes. Contestability: Are challenge path-ways accessible and effective? Harm tracking: Are harms detected, documented, and addressed? Agency protection: Do people have meaningful understanding and choice? Consis-tency: Are ethical decisions coherent across similar cases? Legitimacy: Do stakeholders perceive the system as respectful and accountable?

I. Conclusion

Ethics Without Ideology provides an ethically serious, governance-friendly path for AI in plural societies. By focusing on harms, agency, dignity, and transparency, institutions can build legitimacy without enforcing doctrine. Grounded in Atmakosh-aligned principles of humility and coherence, this approach enables ethical governance that travels across cultures while remaining accountable to human well-being.

REFERENCES

[1] IEEE, “Ethically Aligned Design: A Vision for Prioritizing Human

Well-being with Autonomous and Intelligent Systems,” 1st ed., IEEE Standards Association, 2019. [2] OECD, “OECD Principles on Artificial Intelligence,” Organisation for

Economic Co-operation and Development, 2019. [3] UNESCO, “Recommendation on the Ethics of Artificial Intelligence,”

United Nations Educational, Scientific and Cultural Organization, 2021. [4] NIST, “Artificial Intelligence Risk Management Framework (AI RMF

1.0),” National Institute of Standards and Technology, 2023. [5] ISO/IEC, “ISO/IEC 23894: Information technology — Artificial intel-

ligence — Guidance on risk management,” International Organization for Standardization / International Electrotechnical Commission, 2023.