What is ISO 24028?

ISO 24028, formally titled "Artificial Intelligence — Overview of Trustworthiness in AI Systems," is an international standard published by the International Organization for Standardization (ISO). It provides a framework for understanding and addressing the trustworthiness of artificial intelligence (AI) systems. This standard is designed for developers, organizations, and regulators to ensure AI systems are reliable, ethical, and safe, aligning technological advancements with public trust.

Background and History of ISO 24028

The development of ISO 24028 reflects the increasing need to address the ethical and practical concerns surrounding AI technologies. As AI systems became more advanced and integrated into critical aspects of society, issues like ethical considerations, unintended biases, privacy breaches, and security vulnerabilities emerged as major challenges. These concerns spurred the formation of global initiatives to establish trustworthy AI standards.

ISO began working on ISO 24028 in the early 2020s through the ISO/IEC JTC 1/SC 42 committee, which focuses on AI and related technologies. The standard was published in 2023 as part of an ongoing effort to create a suite of AI-focused standards. Its primary objective is to define "trustworthiness" in AI systems and offer practical guidelines to achieve it. The standard was developed in response to the rapid adoption of AI in diverse fields, ensuring that ethical and safety considerations are integral to technological progress.

Contents of ISO 24028

1. Core Principles of Trustworthy AI ISO 24028 outlines fundamental principles essential for building trustworthy AI systems, including transparency, fairness, accountability, privacy, and security. Transparency ensures AI decisions and models are clearly documented and explainable. Fairness focuses on identifying and mitigating biases in data and algorithms to avoid discriminatory outcomes. Accountability establishes clear ownership and responsibility for AI systems and their impacts. Privacy emphasizes robust data protection mechanisms to safeguard user information, while security prioritizes defending AI systems against malicious attacks or manipulation.
2. Evaluation of Trustworthiness The standard provides a detailed framework for evaluating the trustworthiness of AI systems. It includes metrics to assess various aspects of AI reliability, techniques for validating and verifying AI behavior, and methodologies for auditing AI systems to ensure adherence to the principles outlined.
3. Best Practices for AI Development ISO 24028 offers guidelines for designing and developing AI systems with ethical considerations embedded throughout the lifecycle. These practices focus on ensuring trustworthiness is a foundational component of AI systems, from initial design to deployment and ongoing monitoring.
4. Implementation Roadmap The standard includes step-by-step strategies to help organizations incorporate trustworthiness into their AI projects. It also provides recommendations for complying with ethical and regulatory standards, ensuring alignment with both internal goals and external requirements.

Relevance of ISO 24028

ISO 24028 addresses critical challenges in AI adoption across industries and geographies. For businesses, the standard enables the creation of AI systems that customers and stakeholders can trust, driving broader acceptance and adoption. Governments and regulatory bodies can use it as a benchmark to guide ethical AI development and enforce compliance. On a global scale, ISO 24028 fosters international collaboration, promoting consistency and standardization in AI development practices. Its relevance spans multiple sectors, ensuring that AI technologies benefit society while minimizing risks.

Challenges in Implementing ISO 24028

1. Complexity Implementing ISO 24028 can be complex due to the multifaceted nature of AI technologies and the need for expertise in areas such as ethics, compliance, and AI-specific methodologies.
2. Resources Smaller organizations may face difficulties in allocating the necessary resources, both in terms of finances and skilled personnel, to fully comply with the standard.
3. Dynamic Nature of AI The rapid evolution of AI technologies presents a challenge, as the guidelines established in ISO 24028 must continually adapt to stay relevant.
4. Cultural Variability Trustworthiness in AI is influenced by cultural and societal norms, making it challenging to establish universal definitions and practices that work across all contexts.

Benefits of ISO 24028

1. Enhanced Trust ISO 24028 promotes public confidence in AI systems by embedding principles such as transparency, fairness, and accountability into their design and operation.
2. Risk Mitigation By addressing potential issues such as ethical lapses and unsafe practices, the standard helps organizations reduce risks associated with AI systems.
3. Market Advantage Compliance with ISO 24028 allows organizations to differentiate themselves in the marketplace by showcasing their commitment to trustworthy and ethical AI practices.
4. Regulatory Alignment Organizations adopting the standard are better prepared to meet current and future regulatory requirements, ensuring smooth operations within their respective industries.

Compliance with ISO 24028

1. Gap Analysis Organizations should begin by conducting a thorough gap analysis to compare their existing AI systems with the guidelines outlined in ISO 24028. This process identifies areas where improvements are needed to align with the standard.
2. Training and Awareness Educating employees about the principles and requirements of ISO 24028 is critical. Building a culture of ethical AI development ensures that the standard's principles are integrated into day-to-day practices.
3. System Design Trustworthiness principles should be embedded in the design phase of AI projects. This includes ensuring continual monitoring and validation to maintain compliance throughout the lifecycle of the AI system.
4. Documentation and Reporting Comprehensive documentation should be maintained to demonstrate compliance. This serves as evidence during internal audits and external evaluations.
5. Third-Party Certification Organizations can enhance credibility by engaging accredited bodies to certify compliance with ISO 24028. This external validation reassures stakeholders of the organization's commitment to trustworthy AI.

Conclusion

ISO 24028 is a cornerstone in the effort to create ethical and trustworthy AI systems. By providing a clear framework for evaluating, developing, and maintaining AI systems, it addresses key concerns about transparency, fairness, privacy, and accountability. While implementation can be challenging, the benefits of compliance, including enhanced trust, reduced risks, and market differentiation, make it an essential standard for organizations looking to leverage AI responsibly. Adopting ISO 24028 is not just a regulatory advantage—it is a commitment to building AI systems that positively impact society and inspire confidence among users and stakeholders.

https://www.iso.org/standard/77608.html?

-

#enterpriseriskguy

Muema Lombe, risk management for high-growth technology companies, with over 10,000 hours of specialized expertise in navigating the complex risk landscapes of pre- and post-IPO unicorns.? His new book is out now, The Ultimate Startup Dictionary: Demystify Complex Startup Terms and Communicate Like a Pro?