Towards transparent and explainable AI: Workshop to inform ISO/IEC standards development

Description

You can now view the write-up of the workshop’s output and submit your comments here. 

Transparency and explainability are key requirements for the realisation of trustworthy AI. Building on our webinar on the current standardisation landscape for AI transparency and explainability, this workshop provided an opportunity to help shape two prominent international standards on these topics being developed in ISO/IEC:

• ISO/IEC AWI TS 6254 Information technology — Artificial intelligence — Objectives and approaches for explainability of ML models and AI systems
• ISO/IEC AWI 12792 Information technology — Artificial intelligence — Transparency taxonomy of AI systems

The in-person event was dedicated to gathering input from members of the AI Standards Hub community on selected issues that these standards seek to address and that are currently under discussion at the working group level. Participants received preparatory reading in advance of the workshop to enable active engagement with the issues that were covered during the session. Discussion outcomes were captured by the workshop facilitators and shared to inform the work of the ISO/IEC working group for these standards (ISO/IEC JTC 1/SC 42/Working Group 3).

By participating, community members gained first-hand insight into the nature of AI standards development and had a unique opportunity to share their views to shape these standards at an early stage of development.

Given the interactive nature of the workshop, places were limited, with priority given to community members that attended the webinar. The workshop was open to participants from all stakeholder groups and no prior standards development experience was required.