The AI Standards Hub recently passed its one-year anniversary. To mark the occasion, we look back on some of the highlights from the Hub’s first year. 

Launch

The Hub was officially launched on 12 October 2022 with an audience of over 1000 in-person and remote attendees and opening remarks by the then Minister for Tech and the Digital economy, Damien Collins. Next to a demonstration of the Hub’s new online platform, the event introduced the Hub’s mission and planned activities, with representatives from the Department for Digital, Culture Media and Sport (now in the Department of Science, Innovation and Technology) discussing the Hub’s key role in the government’s wider approach to AI. The event also featured two panels on the importance of multi-stakeholder perspectives in AI standardisation and the role of standards in achieving trustworthy AI, with speakers from the Confederation of British Industries, the Competition and Markets Authority, the Financial Conduct Authority, the Institute for the Future of Work, techUK, Digital Catapult, the Medicines and Healthcare products Regulatory Agency, and the Centre for Data Ethics and Innovation.

A full recording of the launch event is available here. 

Growing the Hub’s AI Standards Database and online community

The Hub’s online platform has attracted a rapidly growing user base, with the most popular resource being the platform’s searchable database for AI standards. During its first year, the content of the database has grown to 354 items, including 212 standards specific to AI and 142 standards concerning broader topics that a relevant to enabling AI.

The last year also saw a steady growth in the number of community members making use of the online platform’s interactive features that come with registering for a user account, with over 1,500 registered community members to date. Popular features include the ability to “follow” standards as a method of bookmarking them for personal reference, receiving updates about their development status, posting comments, and connecting with other users. The database is also designed to highlight standards that are open for comment on the BSI Standards Development Portal, facilitating community input to important standards under development, such as the recently published ISO/IEC FDIS 42001 Artificial Intelligence Management System. 

Live events on trustworthy AI

The Hub hosted various live events dedicated to the broad horizontal theme of ‘trustworthy AI’, including the topics of:

• transparency and explainability
• safety, security and resilience
• uncertainty quantification.

A combination of webinars and workshops, these events aimed to foster a shared understanding of trustworthy AI and its constituent concepts, introduce relevant standardisation efforts, and gather community input to inform the development of selected standards of particular relevance.

Our webinar on ‘European Standardisation in the Context of the EU AI Act’ had nearly 500 live and post-event viewers. This event and all available recordings from the ‘Trustworthy AI’ series can be viewed by following the links below:

• European Standardisation in the Context of the EU AI Act
• Towards Transparent and Explainable AI: The Current Standardisation Landscape
• Enabling Trustworthy AI: Key Developments in International Standardisation
• Measuring Trustworthiness in AI: Approaches and Considerations from a Metrology Perspective

Training offerings

As part of its ‘Knowledge and Training’ pillar, the Hub published seven e-learning modules, alongside the introductory ‘Standards at a Glance’ section and live training events such as our workshop on key concepts for transparent and explainable AI.

Popular e-learning modules include:

• Introduction to Standards
• Introduction to AI Assurance
• Machine learning for Metrology
• Assessing and Mitigating Bias and Discrimination in AI

A full list of available e-learning modules can be accessed here. 

Research and insights

The Hub also published various blogs by experts from The Alan Turing Institute, the British Standards Institution, the National Physical Laboratory, and invited guests. Topics ranged from introducing specific standards, information on the wider landscape, and insights on relevant regulatory developments. You can explore our 5 most popular blogs from the last year here:

• ISO/IEC 23894 – A new standard for risk management of AI
• How ISO/IEC 42001 guides organisations toward Trustworthy AI
• How a standard in development (ISO/IEC 42001) can meet collective AI Governance goals
• What role do standards play in the EU AI Act?
• International trends in AI governance

Informing the relationship between standards and regulation

Finally, the Hub established the AI Standards Forum for UK Regulators, a working group for representatives from the UK’s wide range of regulatory bodies to come together and consider the relationship between the evolving AI standards landscape and regulatory efforts related to AI. Featuring a combination of expert input and peer discussions, the Forum aims to enable knowledge sharing on the multifaceted ways in which standards may support regulatory objectives. Meeting on a regular basis, the Forum has proven highly popular, and currently reaches over 30 regulatory bodies across a diverse range of remits. 

Looking ahead

The AI Standards Hub passed its one-year mark with the publication of an independent evaluation commissioned by the Department for Science, Innovation, and Technology which concluded that its activities thus far have succeeded in establishing it as a key initiative globally. The evaluation established that the Hub has made a strong contribution to raising awareness and building capacity around AI standards in the UK, and has attracted significant attention from around the world, seeing it as well positioned to facilitate novel forms of international collaboration in this space.

Looking forward, the Hub will be guided by these findings and will seek to maintain existing capacity while broadening its focus internationally, through building relationships with multilateral organisations and potential partners in other countries. As a first step in pursuing this increased international engagement, the AI Standards Hub recently participated in the Internet Governance Forum’s annual meeting in Kyoto with a session on the importance of international cooperation and multi-stakeholder inclusion on AI standards, available to watch here. 

Get involved

To continue engaging with the Hub as it enters its next phase, you can:

• Register for a user account and browse our databases for standards, policy, and research, explore our online training materials, and engage in forum discussions.
• Sign up for the Hub’s newsletter and follow on Twitter/X and LinkedIn to stay up to date on events, articles and updates.