Work begins on EU Code of Conduct on GPAI models

(ANSA) - ROME, OCT 4 - The process of developing the EU's first Code of Conduct has officially begun, which will detail the rules of the EU's AI ;;Act for providers of general purpose artificial intelligence (GPAI) models, including those with systemic risks.
    The kick-off plenary session convened by the European AI Office was held, attended by almost a thousand participants, including providers of general purpose AI models, downstream providers, representatives of industry, civil society, academia and independent experts.
    The online meeting, of an operational nature, was open only to interested parties who registered by 25 August 2024.
    Before their publication in the autumn, the AI ;;Office will also present the first results of the stakeholder consultation on the Code of Conduct, on which almost 430 contributions were received. The Code of Conduct aims to facilitate the correct application of the rules set out in the AI ;;Act in relation to general-purpose AI models, including those on transparency and copyright, the systemic risk taxonomy, risk assessment and mitigation measures.
    The process of developing the Code of Conduct involves four working groups meeting three times to discuss drafts. This process will be led by chairs and vice-chairs, including Marta Ziosi, a post-doctoral researcher at the Oxford Martin AI Governance Initiative, and Daniel Privitera, founder and executive director of the Kira Center, an independent non-profit organization based in Berlin that deals with AI. The two Italians will be vice-chairs of the working group 'risk identification and assessment' and the 'technical risk mitigation' respectively. The final version of the Code of Conduct will be published and presented in a closing plenary, scheduled for April 2025.
    photo: OpenAI CEO Sam Altman (ANSA).