Welcome to the latest edition of our newsletter. Read on to find out what has been happening with the AI Standards Hub over the past few weeks.

Welcome to our regular newsletter. Read on to find out what has been happening and what is coming up with the AI Standards Hub.

Header image with portraits of Hub members and title of podcast feature
Call for Views on the Cybersecurity of AI

The cybersecurity of AI is an essential precondition of safety as it ensures (amongst other things) privacy, reliability, and the secure use of models. DSIT is seeking feedback on their proposed approach to AI cybersecurity in a Call for Views on the Cybersecurity of AI. This Call seeks feedback on the proposed approach to AI cybersecurity, including a Code of Practice, and the development of a global standard which sets out baseline security requirements for AI models and systems. The last date to submit your answers to the Call for Views is 9 August 2024.

Access the Call for Views
The EU flag against a blurred backdrop
EU AI Act – Codes of Practice to be developed for GPAI model providers

With the final text of the EU AI Act having been published on 12 July, its implementation is set to be rolled out in phases over the next two years. The Act will become fully enforceable in July 2026, once the harmonised European Standards requested by the European Commission have been written and agreed. The Commission’s standardisation request does not extend to standards for general purpose AI (GPAI) model standards. Instead, Codes of Practice will be developed by the EU AI Office in collaboration with providers of GPAIs and supported by other stakeholder groups to provide an alternative compliance mechanism for GPAI model providers. The Codes of Practice must be developed within nine months of the Act’s first coming into force in August.

Read more about the Codes of Practice
The word AI spelt out against the image of an integrated circuit
King’s Speech 2024 lends insight into new government's AI strategy

Earlier this month, following the recent General Election, Parliament heard the first King’s Speech delivered under the new Labour Government. AI was referenced in the prime minister’s introduction and in the Speech itself, providing insights into the new government's early thinking on AI governance. In the prime minister’s introduction, Keir Starmer announced that “we will harness the power of artificial intelligence as we look to strengthen safety networks”, while the Speech itself pointed toward “appropriate legislation to place requirements on those working to develop the most powerful artificial intelligence models”. These statements indicate an intention to prioritise and capitalise on AI’s innovative potential while introducing constraints on the most powerful AI systems.

Read the King’s Speech 2024
New blogpost: ISO/IEC 42001: your starting point with AI implementation

In this blog, BSI’s Nathan Parker explains the importance of ISO/IEC 42001, an AI management system standard. The blog provides insights about what the standard can offer to businesses and shares implementation resources for the standard.

Read the blog
multi-coloured puzzle pieces stacked together
Image of Earth from above featuring network connections
New blogpost: Updates from our AI in recruitment and employment workshop series: Refining the scope of future work

In this blog, we share high-level findings emerging from a workshop series we held in February and March on the role of SDO standards in governing the use of AI in recruitment and employment. We provide an overview of key workshop aims and outcomes, explaining how these outcomes will enable us to refine the scope of future work in this area.

Read the blog
orange graphic featuring graduation cap with thought bubble and pathways leading away
Training module of the month - Assessing and mitigating bias and discrimination in AI

This course explains what is meant by bias in the context of machine learning algorithms, how bias can be present in the training data, and how it can be unintentionally introduced in the learning phase. Learners will gain an appreciation of why this is a concern and why it needs to be addressed.

Explore Assessing and mitigating bias and discrimination in AI
Sign up for a user account

Creating a user account on the AI Standards Hub website will enable you to take advantage of a wide range of interactive features. Benefits include the ability to save standards and other database items to your personal dashboard, notifications when followed standards open for comment or move from draft to publication, being able to contribute to our community discussion forums, access to our e-learning modules, and much more!

Set up a user account
An icon of a group of people having a conversation.

Was this newsletter forwarded to you by someone else? Subscribe here to receive future editions directly in your inbox.

Unsubscribe   |   Manage your subscription   |   View online