ISO's aim for standardization in machine learning
summary
Claims that internationally recognized ISO management system standards can make machine learning adoption trustworthy and auditable, offering concrete guidance rather than vague best practices. Cites ISO’s scale — about 24,000 standards and 165 member bodies — and concrete moves like a repository of use cases and domain guidelines for autonomous driving and manufacturing. Flags real problems (poor data quality, interpretability, infrastructure) and a concrete harm: bias in pulse oximeters leads to wrong triage and fatalities. Notes demand signals such as ISO 27001 interest increasing tenfold.
AI-written summary · about
chapters (5)
- 0:00Introduction to ISO ML standardization
- 5:17Why standardization matters
- 13:31ISO standards for AI ethics
- 16:40Management systems adoption
- 19:49Q&A on ISO standards
related talks
31mAI Trustworthiness: Path to Mass Adoption in Critical Industries
33mOpen Standards for Machine Learning Deployment
01h 07mNew Governance and Inclusiveness in AI Standardisation How Far Have We Gone?
23mDefining Open Source AI: Can the “Judgement of Solomon” Help the Open Source Community Find Success?
01h 10mHOW CAN REGULATION HELP BUILD TRUSTWORTHY ARTIFICIAL INTELLIGENCE?
57mWriting Safety Critical Automotive Software for High Perf AI Hardware
55mAttacking And Protecting Artificial Intelligence
20mResponsible AI: Policy Solutions for DevSecOps