The European AI Office recently launched a Multi-stakeholder Consultation on trustworthy general-purpose AI models under the AI Act. Submissions must be completed by Tuesday 10 September 2024.

This will inform the drafting process to detail rules for general-purpose AI model providers in the first General-Purpose AI Code of Practice. It will also inform the AI Office’s related work for a template of the summary content used for the training of those models. This initiative came just a few days before the AI Act came into force on August 1.

The office wants to ensure that you all have the opportunity to participate.

In addition, you can also express your interest by Sunday 25 August 2024 to participate in the 9-month inclusive and transparent drafting process for the first General-Purpose AI Code of Practice between September 2024 and April 2025.

A Code of Practice Plenary with Working Groups will be established, consisting of all interested and eligible general-purpose AI model providers and stakeholders. For each of the four Working Groups of the Plenary, Chairs and, as appropriate, Vice-Chairs are responsible for synthesising submissions from the consultation and Plenary participants. One chair will be designated to take on a central coordination role. Interested independent experts can apply for the role of Working Group Chairs and Vice-Chairs.

For both initiatives, a wide range of stakeholders — including general-purpose AI model providers, downstream providers that integrate a model into their AI system, other industry organisations, civil society, rights holders, academic experts, and others — are invited to contribute to the first draft of the Code of Practice and related tasks. This covers topics such as transparency, copyright-related rules, risk assessment and mitigation in light of systemic risks.

It would be greatly appreciated if these initiatives could also be shared within respective networks.