24 June 2025 – A new global framework to help evaluate how AI systems affect individuals, communities and society has been published by BSI, in recognition of the growing power of AI and especially generative AI tools across all aspects of life, from medicine to retail and public services.
Information technology – Artificial intelligence – AI system impact assessment (BS ISO/IEC 42005:2025) is the world’s first internationally recognized standard for assessing the societal, ethical and operational impact of AI systems.
As AI adoption surges, regulatory scrutiny is intensifying. The introduction of the EU AI Act and growing public concern over AI incidents highlight the urgent need for consistent transparent assessment tools. The new standard, published by BSI in its capacity as the UK national standards body, offers organizations a structured globally accepted framework to demonstrate accountability, which complies with emerging laws and reduces reputational and legal risks.
Unlike general AI governance frameworks, BS ISO/IEC 42005 stands out as the first international standard to deliver a detailed step-by-step methodology for evaluating the impact of AI systems — both within organizations and in their external applications. It equips users to assess foreseeable uses of AI, analyze potential direct and indirect effects on individuals and society, and integrate these assessments into broader AI management systems like BS ISO/IEC 42001.
Developed with input from a global network of experts, the standard places human impact at its core and ensures ethical considerations are not just theoretical, but practical, whether it is a healthcare provider addressing diagnostic bias or a public authority scrutinizing algorithmic decision-making.
This standard supports a broad spectrum of stakeholders, including technology and compliance leaders, AI developers and integrators, policymakers, regulators, auditors, SMEs, and manufacturers creating AI-powered solutions.
David Cuckow, Director of Digital, BSI said “AI is starting to shape all aspects of society, from healthcare to public service delivery and beyond, but people and businesses need confidence about its societal, ethical and operational impacts. The new standard BS ISO/IEC 42005 marks a pivotal step towards trustworthy AI, by putting people’s rights, dignity, and well-being at the core of AI assessment. For the first time, organizations have a globally endorsed roadmap for understanding and mitigating the real-world impacts of AI systems and will be empowered to align innovation with accountability to ensure that as AI evolves, so too does our commitment to ethical, human-centred development.”
For further information about the standard, please visit https://knowledge.bsigroup.com/products/information-technology-artificial-intelligence-ai-ai-system-impact-assessment
