AI Standards Bodies

Overview

AI standards bodies represent one of the most influential yet under-examined mechanisms shaping AI governance worldwide. These organizations develop technical specifications that, while typically voluntary, create powerful incentives for compliance through regulatory integration, procurement requirements, and industry coordination. Unlike direct regulation, standards operate through market mechanisms and professional norms, making them particularly effective at scaling governance practices across jurisdictions and sectors.

The strategic importance of AI standards has become increasingly evident as major regulatory frameworks like the EU AI ActPolicyEU AI ActComprehensive overview of the EU AI Act's risk-based regulatory framework, particularly its two-tier approach to foundation models that distinguishes between standard and systemic risk AI systems. ...Quality: 55/100 explicitly incorporate them as compliance pathways. When regulations reference specific standards, following those standards creates a "presumption of conformity" - essentially a safe harbor that reduces legal risk. This regulatory integration transforms voluntary technical documents into de facto requirements for companies operating in multiple markets. Understanding how standards bodies operate, which organizations hold influence, and how safety considerations are embedded in their processes has become essential for anyone working on AI governance.

Key organizations include ISO/IEC JTC 1/SC 42 for international standards, IEEE for technical specifications, and regional bodies like CEN-CENELEC developing EU AI Act compliance standards. Current standards address risk management (ISO/IEC 23894), management systems (ISO/IEC 42001), and ethical considerations (IEEE 7000 series), with harmonized EU standards expected 2025-2026 to provide direct compliance pathways for high-stakes AI applications.

Major AI Standards Organizations Comparison

International Standards Architecture

ISO/IEC Joint Technical Committee

ISO/IEC JTC 1/SC 42 serves as the primary international forum for AI standardization, bringing together the International Organization for Standardization and International Electrotechnical Commission under a joint technical committee structure established in 2017. This body has produced the most comprehensive suite of AI standards currently available, with over 30 published standards and technical reports covering everything from basic terminology to complex risk management frameworks.

The committee's flagship achievement is ISO/IEC 23894 on AI Risk Management, published in 2023, which provides a systematic framework for identifying, assessing, and mitigating risks throughout the AI lifecycle. This standard has been explicitly referenced in the EU AI Act and NIST AI Risk Management FrameworkPolicyNIST AI Risk Management Framework (AI RMF)Comprehensive analysis of NIST AI RMF showing 40-60% Fortune 500 adoption with implementation costs of $50K-$1M+ annually, but lacking quantitative evidence of actual risk reduction and inadequate ...Quality: 60/100, demonstrating how international standards can achieve global influence through regulatory adoption. The standard establishes risk categories including accuracy, security, transparency, and societal impact, with specific guidance on risk assessment methodologies and documentation requirements.

ISO/IEC 42001, the AI Management System standard published in 2023, takes a different approach by establishing organizational requirements for responsible AI development and deployment. Modeled on successful management system standards like ISO 9001 for quality and ISO 27001 for information security, this standard enables third-party certification of an organization's AI governance capabilities. Early adopters include major technology companies and consulting firms seeking to demonstrate systematic approaches to AI risk management to clients and regulators.

The committee's work on terminology (ISO/IEC 22989) has proven surprisingly influential by establishing standardized definitions for concepts like "AI system," "machine learning," and "trustworthiness." These definitions are being incorporated into regulatory frameworks worldwide, creating consistency in how AI systems are categorized and evaluated. This seemingly mundane standardization work has significant implications for which systems fall under specific regulatory requirements.

Key Published ISO/IEC AI Standards

As of December 2024, ISO/IEC 42001 certification↗🔗 webISO/IEC 42001 certificationRelevant as an example of corporate AI governance certification efforts; ISO/IEC 42001 is an emerging international standard for AI management systems that may inform regulatory and industry compliance frameworks.Cognizant announced it became the first company to achieve accredited certification under ISO/IEC 42001:2023, the international standard for Artificial Intelligence Management S...governancepolicydeploymentai-safety+1Source ↗ has been achieved by organizations including KPMG Australia, Cognizant, Microsoft (for M365 Copilot), OrionStar Robotics, and Synthesia, demonstrating growing market adoption of systematic AI governance approaches.

IEEE Standards Association

The Institute of Electrical and Electronics Engineers has approached AI standardization through its established ethics-focused standards program, producing the IEEE 7000 series that addresses ethical design processes rather than just technical specifications. IEEE 7000, published in 2021, establishes a model process for addressing ethical concerns during system development, requiring organizations to identify stakeholders, analyze potential harms, and implement mitigation measures throughout the design process.

IEEE 7001 on transparency of autonomous systems has gained particular attention for its practical guidance on explainable AI requirements. The standard provides specific metrics and testing procedures for evaluating whether AI systems provide sufficient transparency for their intended use contexts. This work has influenced regulatory discussions about explainability requirements, particularly for high-risk applications like healthcare and criminal justice.

The IEEE Ethics Certification Program for Autonomous Systems (ECPAIS) represents an innovative approach to standards implementation by offering third-party certification based on the IEEE 7000 series. Organizations can obtain certification by demonstrating compliance with ethical design processes and transparency requirements through independent audits. While still in early stages, this certification program could become a market differentiator as organizations seek to demonstrate responsible AI practices to stakeholders.

IEEE's collaborative approach involves extensive consultation with civil society organizations, academic institutions, and professional associations beyond the technology industry. This broader stakeholder engagement has resulted in standards that address societal impacts more comprehensively than purely technical specifications, though it has also led to longer development timelines and more complex implementation requirements. According to research published in Frontiers in Robotics and AI↗🔗 webresearch published in Frontiers in Robotics and AIA peer-reviewed IEEE standardization proposal offering a concrete, stakeholder-differentiated framework for transparency in AI and robotics; relevant to governance and accountability discussions in AI safety.This paper presents IEEE P7001, a draft standard that operationalizes transparency in autonomous systems as a measurable, testable property rather than a vague ethical principle...governancepolicyinterpretabilityevaluation+3Source ↗, transparency appears in 87% of AI ethics guidelines surveyed (73 of 84 sets), making it the most frequently included ethical principle.

IEEE 7000 Series Standards

Regional Standards Development

European Harmonized Standards

The European Committee for Standardization (CEN) and European Committee for Electrotechnical Standardization (CENELEC) have been tasked with developing harmonized standards that provide presumption of conformity with the EU AI Act, representing the most direct integration of standards and regulation in AI governance to date. The European Commission issued formal standardization requests in May 2023, with over 1,000 European experts from more than 20 countries↗🔗 webCEN-CENELEC JTC 21: European AI Standardization CommitteeRelevant for those tracking EU AI governance and regulatory compliance; JTC 21's harmonized standards will directly shape how high-risk AI systems must be designed and assessed under the EU AI Act.CEN-CENELEC Joint Technical Committee 21 (JTC 21) is the European standardization body developing harmonized AI standards to support the EU AI Act. With over 300 experts across ...governancepolicyai-safetydeployment+3Source ↗ now participating in this effort - the largest coordinated AI standardization initiative in European history.

These harmonized standards will address specific requirements for high-risk AI systems, including risk assessment methodologies, quality management systems, data governance requirements, and human oversight mechanisms. Following these standards will create a legal presumption that AI systems comply with EU AI Act requirements, providing companies with clear compliance pathways and reducing regulatory uncertainty. This approach leverages the EU's single market power to influence global AI standards, as companies developing AI systems for multiple markets often adopt the most stringent requirements as a baseline.

EU AI Act Harmonized Standards Timeline

The European Telecommunications Standards Institute (ETSI) has complemented this work through its Securing Artificial Intelligence (SAI)↗🔗 webETSI Technical Committee on Securing Artificial Intelligence (TC SAI)ETSI TC SAI is a key international standards body for AI security; its specifications are relevant to AI governance discussions and technical safety practitioners concerned with deployment-time robustness and regulatory compliance.ETSI's Technical Committee on Securing Artificial Intelligence (TC SAI) develops global standards to address AI security threats, including adversarial attacks, data poisoning, ...governancepolicytechnical-safetydeployment+4Source ↗ series, focusing on cybersecurity aspects of AI systems. ETSI's standards address adversarial attacks, data poisoning, model theft, and other security vulnerabilities that could compromise AI system performance or enable malicious use. The flagship ETSI TS 104 223↗🔗 webETSI TS 104 223: Baseline Cybersecurity Requirements for AI Models and SystemsAn official ETSI technical standard relevant to AI safety practitioners and policymakers; provides concrete normative cybersecurity requirements that may inform regulatory compliance and AI deployment best practices in Europe and internationally.ETSI TS 104 223 is a formal technical standard published in April 2025 that establishes baseline cybersecurity requirements for AI models and systems. It organizes security guid...governancepolicytechnical-safetydeployment+3Source ↗ defines 13 core principles expanding to 72 trackable requirements across 5 lifecycle phases. This security focus reflects European concerns about AI systems' potential vulnerabilities to state-sponsored attacks and criminal exploitation, with direct relevance to the Cyber Resilience Act and NIS2 Directive.

CEN-CENELEC's Joint Technical Committee 21 on AI has established working groups addressing conformity assessment procedures, testing methodologies, and certification requirements. These groups are developing practical guidance for how third-party assessment bodies should evaluate AI systems for EU AI Act compliance, including specific testing protocols and documentation requirements that will shape how AI systems are validated across Europe.

National Standards Initiatives

The National Institute of Standards and Technology (NIST) in the United States has taken a framework-based approach rather than developing formal standards, producing the AI Risk Management Framework (AI RMF 1.0)↗🏛️ government★★★★★NISTNIST AI Risk Management FrameworkThe NIST AI RMF is a widely referenced U.S. government standard for AI risk governance, frequently cited in policy discussions and used by organizations building internal AI safety and compliance programs; relevant to AI safety researchers tracking institutional governance approaches.The NIST AI RMF is a voluntary, consensus-driven framework released in January 2023 to help organizations identify, assess, and manage risks associated with AI systems while pro...governancepolicyai-safetydeployment+4Source ↗ released on January 26, 2023 as directed by the National Artificial Intelligence Initiative Act of 2020. NIST's approach emphasizes voluntary adoption and industry self-regulation while providing detailed guidance on risk assessment, mitigation strategies, and governance processes through four core functions: Govern, Map, Measure, and Manage.

NIST's Generative AI Profile (AI 600-1)↗🏛️ government★★★★★NISTGenerative AI Profile (AI 600-1)An official U.S. government framework document from NIST; highly relevant for organizations seeking compliance guidance or structured risk management approaches for generative AI, and a key reference for AI governance discussions in policy and industry contexts.NIST AI 600-1 is a specialized profile of the AI Risk Management Framework (AI RMF 1.0) focused on generative AI systems, identifying 12 risk categories unique to or exacerbated...governancepolicyai-safetydeployment+3Source ↗, published in 2024, addresses specific risks associated with large language modelsCapabilityLarge Language ModelsComprehensive analysis of LLM capabilities showing rapid progress from GPT-2 (1.5B parameters, 2019) to GPT-5 and Gemini 2.5 (2025), with training costs growing 2.4x annually and projected to excee...Quality: 60/100 and other generative AI systems. This profile identifies unique risks including hallucinations, content provenance issues, and potential for misuse in disinformationRiskAI DisinformationPost-2024 analysis shows AI disinformation had limited immediate electoral impact (cheap fakes used 7x more than AI content), but creates concerning long-term epistemic erosion with 82% higher beli...Quality: 54/100 campaigns. The profile's influence extends beyond the US through adoption by multinational companies and integration into procurement requirements by other governments. The AI RMF is designed as a "living document" with formal community review expected no later than 2028.

The British Standards Institution (BSI) has focused on ethical considerations through standards like BS 8611 on ethical design for robotics and AI systems. BSI's approach emphasizes stakeholder engagement and impact assessment throughout the development lifecycle, reflecting UK policy priorities around responsible innovation and public engagement with emerging technologies.

Standards Australia has developed a comprehensive AI governance framework that adapts international standards to Australian regulatory and cultural contexts. This localization approach demonstrates how national standards bodies can leverage international work while addressing specific domestic priorities around data sovereignty, indigenous rights, and regional economic development.

Standards Ecosystem Architecture

The following diagram illustrates how international and regional standards bodies interact with regulatory frameworks to create compliance pathways:

flowchart TD
  subgraph International["International Standards"]
      ISO["ISO/IEC JTC 1/SC 42"]
      IEEE["IEEE Standards Association"]
  end

  subgraph Regional["Regional Bodies"]
      CEN["CEN-CENELEC JTC 21"]
      ETSI["ETSI TC SAI"]
      NIST["NIST"]
  end

  subgraph Regulations["Regulatory Frameworks"]
      EUAI["EU AI Act"]
      USEO["US Executive Order"]
      UKAIA["UK AI Framework"]
  end

  subgraph Outcomes["Compliance Mechanisms"]
      CERT["Third-Party Certification"]
      PROC["Procurement Requirements"]
      PRES["Presumption of Conformity"]
  end

  ISO --> CEN
  ISO --> CERT
  IEEE --> CERT
  CEN --> EUAI
  ETSI --> EUAI
  NIST --> USEO
  EUAI --> PRES
  USEO --> PROC
  CERT --> PROC

Standards Implementation and Market Dynamics

Certification and Compliance Pathways

The emergence of third-party certification programs based on AI standards represents a significant development in how organizations demonstrate responsible AI practices. ISO/IEC 42001 certification, offered by major certification bodies like BSI, SGS, and Bureau Veritas, requires organizations to implement comprehensive AI governance systems including risk assessment procedures, stakeholder engagement processes, and continual improvement mechanisms.

Certification processes typically involve initial gap assessments, implementation support, and formal audits conducted by trained assessors. Organizations must demonstrate not just policy compliance but effective implementation through documented evidence of risk assessments, stakeholder consultations, and incident response procedures. This rigorous assessment process has driven substantive improvements in AI governance practices among early adopters.

The business case for standards compliance has strengthened significantly as procurement requirements increasingly reference specific standards. The US federal government has begun requiring NIST framework compliance for AI systems used in federal agencies, while European public sector procurement increasingly references ISO standards for AI systems. These procurement requirements create market incentives for standards adoption that extend beyond regulatory compliance.

Insurance companies have also begun incorporating AI standards compliance into coverage decisions and premium calculations. Organizations demonstrating compliance with recognized standards may qualify for reduced premiums or expanded coverage for AI-related liabilities, creating additional market incentives for standards adoption.

According to ANAB (ANSI National Accreditation Board)↗🔗 webANAB (ANSI National Accreditation Board)Relevant for those tracking formal AI governance standards and certification frameworks; ISO/IEC 42001 is increasingly referenced in regulatory and procurement contexts as a benchmark for organizational AI risk management.This resource from ANAB (ANSI National Accreditation Board) explains the ISO/IEC 42001 standard for AI Management Systems, which provides a framework for organizations to respon...governancepolicydeploymentai-safety+2Source ↗, demand for ISO/IEC 42001 certification has been substantial, with 15 certification bodies applying for accreditation by late 2024. Eurostat data↗🔗 webISO 42001:2023 AI Management System Certification GuideUseful for those researching AI governance standards and certification frameworks; ISO 42001:2023 represents a key institutional mechanism for operationalizing responsible AI practices in organizational settings.This resource from Certiget provides a guide to ISO 42001:2023, the international standard for AI management systems, explaining its certification aims, scope, and relevance for...governancepolicydeploymentevaluation+2Source ↗ indicates that 13.5% of EU enterprises (with at least 10 employees) used at least one AI technology in 2024, up from 8% in 2023, with 41.2% adoption among large enterprises - creating a rapidly expanding market for AI governance standards.

Industry Influence and Participation Patterns

Major technology companies including Microsoft, Google, IBM, and Amazon actively participate in standards development through dedicated standards teams and executive-level engagement. These companies typically assign senior technical staff to standards committees and provide substantial resources for standards development activities, giving them significant influence over standards content and development priorities.

This industry participation has created concerns about regulatory capture, where standards may reflect industry preferences rather than broader public interests. However, industry engagement has also brought essential technical expertise and implementation experience that has improved standards quality and practical applicability. The challenge lies in balancing industry input with other stakeholder perspectives.

Academic institutions have played important roles in AI standards development, particularly in areas requiring specialized expertise like machine learning robustness, bias assessment, and ethical design methodologies. Universities often provide neutral venues for standards development activities and contribute research-based evidence for standards requirements.

Civil society participation remains limited but growing, with organizations like AI Now Institute, Partnership on AI, and IEEE Society on Social Implications of Technology contributing to standards development processes. These organizations often focus on ensuring standards address societal impacts and marginalized communities' concerns, though they face resource constraints that limit sustained participation.

Safety Implications and Risk Considerations

Opportunities for Safety Integration

AI standards offer several mechanisms for embedding safety considerations into industry practice at scale. Risk management standards like ISO/IEC 23894 require systematic identification and assessment of potential harms, including safety risks to individuals and society. These requirements create organizational incentives for proactive safety consideration rather than reactive responses to incidents.

Management system standards establish ongoing governance processes that can catch safety issues before they escalate. ISO/IEC 42001 requires regular risk assessments, incident reporting systems, and continual improvement processes that help organizations identify and address safety concerns throughout AI system lifecycles. The certification requirements create external accountability for maintaining these processes.

Technical standards addressing testing, validation, and transparency provide tools for evaluating AI system safety characteristics. IEEE standards on algorithmic bias, system transparency, and robustness testing offer specific methodologies for assessing whether AI systems meet safety requirements. These standards enable more systematic and comparable safety assessments across organizations and applications.

The international nature of standards development creates opportunities for spreading safety practices globally, including to jurisdictions with less developed AI governance frameworks. Companies operating internationally often adopt the most stringent standards as baseline practices, creating a "regulatory ratchet" effect that can improve safety practices worldwide.

Standards Impact Assessment

Limitations and Potential Risks

Standards-based approaches to AI safety face several inherent limitations that could create false assurance or inadequate protection. The consensus-based nature of standards development often produces minimum viable requirements rather than best practices, as standards must accommodate diverse industry perspectives and capabilities. This "lowest common denominator" effect can result in standards that provide appearance of safety without substantive protection.

The voluntary nature of most standards means compliance depends on market incentives rather than legal requirements. Even when standards are incorporated into regulations, enforcement often relies on self-certification or limited oversight rather than comprehensive monitoring. Organizations may achieve technical compliance while missing the underlying safety objectives that standards are intended to support.

Standards development timelines often lag significantly behind technology development, creating gaps where rapidly evolving AI capabilities lack appropriate safety standards. Large language models, multimodal AI systems, and AI agents present novel risks that current standards may not adequately address. The multi-year standards development process cannot easily adapt to the pace of AI advancement.

The technical complexity of AI systems creates challenges for standards implementation and verification. Many AI safety properties are difficult to measure objectively, leading to standards requirements that are subjective or difficult to verify consistently across different assessment bodies and contexts.

Future Trajectory and Strategic Implications

Near-Term Developments (2025-2026)

The completion of EU AI Act harmonized standards represents the most significant near-term development in AI standards, creating the first comprehensive regulatory integration of AI standards worldwide. These standards will establish specific compliance pathways for high-risk AI applications and influence global practices through the Brussels Effect, where EU regulations shape global industry practices.

ISO/IEC standards development will accelerate in response to regulatory demand, with new standards addressing generative AI, AI agents, and sector-specific applications. The success of early management system certifications will likely drive expanded certification programs and more sophisticated third-party assessment capabilities.

NIST framework updates will incorporate lessons learned from initial implementation and address emerging technologies like multimodal AI and automated decision systems. These updates will likely influence international standards development and provide templates for other national approaches to AI governance.

Regional standards bodies will develop localized versions of international standards, addressing specific regulatory requirements, cultural contexts, and economic priorities. This localization trend will create both opportunities for innovation and challenges for multinational companies managing diverse compliance requirements.

Medium-Term Evolution (2-5 years)

AI standards will likely evolve toward more automated compliance assessment and real-time monitoring capabilities. Current standards rely heavily on documentation and periodic assessments, but future standards may incorporate continuous monitoring, automated testing, and algorithmic auditing capabilities that provide ongoing assurance of standards compliance.

Sector-specific standards will emerge for healthcare, finance, transportation, and other domains with specialized AI safety requirements. These standards will address domain-specific risks and regulatory requirements while building on foundational AI governance standards. Professional associations and sector regulators will play larger roles in developing and enforcing these specialized standards.

International coordination mechanisms will strengthen as countries recognize the benefits of harmonized approaches to AI standards. Bilateral and multilateral agreements may establish mutual recognition of standards compliance and certification programs, reducing regulatory fragmentation and compliance costs for multinational organizations.

The relationship between standards and AI safety research will deepen, with standards development incorporating emerging research on AI alignment, robustness, and interpretability. This integration will help translate research insights into practical governance tools while providing feedback on real-world implementation challenges.

Key Uncertainties and Strategic Questions

The effectiveness of current AI standards in addressing existential risks from advanced AI systems remains unclear. Most existing standards focus on near-term applications and incremental improvements in AI governance rather than the fundamental challenges posed by artificial general intelligence or superintelligence. Whether standards-based approaches can scale to address these advanced risks represents a crucial uncertainty.

The balance between industry self-regulation through standards and direct government regulation continues to evolve across jurisdictions. Some governments may conclude that voluntary standards are insufficient for AI governance and pursue more directive regulatory approaches, while others may rely primarily on standards-based frameworks. This variation could create significant compliance complexity and competitive distortions.

The participation of non-Western countries in AI standards development will significantly influence global AI governance. China, India, and other major economies are developing their own AI standards capabilities and may pursue alternative approaches that diverge from current international standards. The degree of convergence or fragmentation in global AI standards will shape the effectiveness of standards-based governance approaches.

The integration of AI standards with emerging technologies like quantum computing, biotechnology, and autonomous systems will create new governance challenges. Current standards may prove inadequate for AI systems that operate in physical environments or interact with other advanced technologies, requiring fundamental rethinking of standards approaches.

For those concerned about AI safety, engaging with standards development processes offers both opportunities and challenges. Standards can embed safety considerations into industry practice at scale, but they can also create false assurance or capture by industry interests. The most effective approach likely involves sustained participation in standards development while maintaining realistic expectations about their limitations and advocating for complementary governance mechanisms.

Key References

• ISO/IEC JTC 1/SC 42 - Artificial Intelligence↗🔗 webISO/IEC JTC 1/SC 42 – Artificial Intelligence Standards CommitteeISO/IEC JTC 1/SC 42 is the premier international AI standards committee; its outputs (e.g., ISO/IEC 42001, 23894) are increasingly referenced in national AI regulations and procurement frameworks globally.ISO/IEC JTC 1/SC 42 is the primary international standards committee responsible for AI standardization, operating under joint ISO/IEC governance with ANSI as secretariat. It de...governancepolicyai-safetydeployment+3Source ↗ - Official ISO committee page
• ISO/IEC 42001:2023 - AI Management Systems↗🔗 webISO/IEC 42001:2023 – AI Management System StandardISO/IEC 42001:2023 is a key international governance standard for AI management systems; relevant to organizations seeking structured frameworks for responsible AI deployment and increasingly referenced in regulatory compliance contexts.ISO/IEC 42001:2023 is the first international standard specifying requirements for establishing, implementing, maintaining, and continually improving an Artificial Intelligence ...governancepolicydeploymentai-safety+3Source ↗ - Management system standard
• ISO/IEC 23894:2023 - AI Risk Management Guidance↗🔗 webISO/IEC 23894:2023 – Information Technology: Artificial Intelligence – Guidance on Risk ManagementThis ISO standard is a key international reference for organizations seeking to formalize AI risk management practices; relevant to policy makers, compliance teams, and AI safety practitioners working on institutional governance frameworks.ISO/IEC 23894:2023 is an international standard providing guidance for organizations that develop, produce, deploy, or use AI systems on how to manage AI-specific risks. It desc...governancepolicyai-safetydeployment+2Source ↗ - Risk management standard
• NIST AI Risk Management Framework↗🏛️ government★★★★★NISTNIST AI Risk Management FrameworkThe NIST AI RMF is a widely referenced U.S. government standard for AI risk governance, frequently cited in policy discussions and used by organizations building internal AI safety and compliance programs; relevant to AI safety researchers tracking institutional governance approaches.The NIST AI RMF is a voluntary, consensus-driven framework released in January 2023 to help organizations identify, assess, and manage risks associated with AI systems while pro...governancepolicyai-safetydeployment+4Source ↗ - US voluntary framework
• IEEE 7000 Series Projects↗🔗 webIEEE 7000-2021: Model Process for Addressing Ethical Concerns During System DesignIEEE 7000-2021 is a formal international standard relevant to AI governance and responsible design; useful for organizations implementing structured ethical review processes in AI/autonomous system development.IEEE 7000-2021 is a technical standard providing a process framework for incorporating ethical values into the design and development of autonomous and AI-enabled systems. It of...governanceai-safetypolicyalignment+3Source ↗ - Ethics-focused standards
• CEN-CENELEC Artificial Intelligence↗🔗 webCEN-CENELEC JTC 21: European AI Standardization CommitteeRelevant for those tracking EU AI governance and regulatory compliance; JTC 21's harmonized standards will directly shape how high-risk AI systems must be designed and assessed under the EU AI Act.CEN-CENELEC Joint Technical Committee 21 (JTC 21) is the European standardization body developing harmonized AI standards to support the EU AI Act. With over 300 experts across ...governancepolicyai-safetydeployment+3Source ↗ - EU harmonized standards
• ETSI Securing Artificial Intelligence↗🔗 webETSI Technical Committee on Securing Artificial Intelligence (TC SAI)ETSI TC SAI is a key international standards body for AI security; its specifications are relevant to AI governance discussions and technical safety practitioners concerned with deployment-time robustness and regulatory compliance.ETSI's Technical Committee on Securing Artificial Intelligence (TC SAI) develops global standards to address AI security threats, including adversarial attacks, data poisoning, ...governancepolicytechnical-safetydeployment+4Source ↗ - AI cybersecurity standards
• EU AI Act Standardisation↗🔗 web★★★★☆European UnionEU AI Act StandardisationOfficial EU Commission page on AI Act standardisation efforts; essential reference for understanding how the EU's binding AI regulation translates into technical compliance standards via CEN/CENELEC, relevant to anyone tracking AI governance and global regulatory benchmarking.This European Commission page explains how harmonised technical standards are being developed under the EU AI Act to translate legal requirements into common technical language....governancepolicyai-safetydeployment+2Source ↗ - European Commission policy page
• Winfield & Jirotka (2021) - IEEE P7001 Transparency Standard↗🔗 webresearch published in Frontiers in Robotics and AIA peer-reviewed IEEE standardization proposal offering a concrete, stakeholder-differentiated framework for transparency in AI and robotics; relevant to governance and accountability discussions in AI safety.This paper presents IEEE P7001, a draft standard that operationalizes transparency in autonomous systems as a measurable, testable property rather than a vague ethical principle...governancepolicyinterpretabilityevaluation+3Source ↗ - Academic analysis of transparency in AI ethics guidelines