MLex: Providers of general-purpose AI (GPAI) models including OpenAI, Google and Microsoft have had confirmation that the legal deadline for a final Code of Practice to help them comply with the EU AI Act will be missed and that they may not see it for a couple more months. The delay is due to efforts to align the Code with guidelines on GPAI rules, and to secure the support of major AI companies amid pressure from the US Government.

MLex: The European Commission adopted an implementing Act setting out the establishment and operation of the scientific panel of independent experts that will advise and support EU and national authorities on the enforcement of the EU AI Act’s rules on general-purpose AI models. The final Act includes significant changes compared to an earlier draft on sensitive matters such as conflict of interest, internal procedures, and transparency.

MLex: AI model providers saw the European Commission publish its preliminary approach to the EU AI Act’s rules for general-purpose AI (GPAI) as part of a public consultation meant to inform upcoming guidelines on this matter. The preliminary approach gives insights into the Commission’s current understanding of key concepts, including setting compute thresholds for determining whether a model is covered under the AI rulebook and when legal duties shift to downstream players. The working document provides examples for when a model is considered placed on the EU market, methodologies for calculating computational resources, and explains how a code of practice that is currently being developed is meant to support compliance with the EU AI Act.

MLex: Companies providing the most powerful AI models are seeing a new version of a code of practice to help them comply with the EU's AI Act. The new rules are less prescriptive and more outcome-oriented. Noteworthy changes include a shortened risk taxonomy, toned-down whistleblower protection, more clarity on when third-party evaluators would be required, and a suggestion to introduce a review mechanism, including for emergency cases.

MLex: A German collective rights management organisation has won its copyright lawsuit against Suno, developer of an AI music generator. A Munich court found that Suno reproduced copyrighted music without permission and granted the collective rights body, GEMA, its requests for damages and injunctive relief in a ruling that could open the door for more European copyright actions against AI companies.

MLex: A photographer's image, which has been downloaded for creating a database to be used for training artificial intelligence (AI) image generators, can be used under an exception in a German law implementing Directive (EU) 2019/790 (the EU Copyright Directive). The ruling by the Hamburg Regional Court handed down on 27 September 2024 could lead to creators, such as the photographer Robert Kneschke, having less control over the use of their work for AI training.

The Department for Science, Innovation & Technology (DSIT), acting on behalf of the Ministry of Housing, Communities and Local Government (MHCLG), has awarded a contract for the development of an AI-augmented planning decisions tool aimed at reducing processing times for planning applications. Five companies tendered, and Google Cloud EMEA has been selected as the supplier. The contract is worth just over £6.9 m and will run until at least May 2028.

MLex: Leading AI companies have been presented with the preliminary findings of expert reports that will inform a voluntary EU code of practice on labelling AI-generated content under the EU AI Act. The studies on text, audio, image and video highlight that no single technique meets legal requirements, recommending multi-layered strategies combining methods. The timing of the code remains unclear.

Law360, Expert Analysis: Between the European Union Artificial Intelligence Act's entry into force on 1 August 2024 and the UK Labour government's announcement in the King's Speech on 17 July 2024 of its intention to introduce new AI reforms, it is clear that regulators in Europe are aiming to set clear guardrails on AI development and deployment. Written by Nicola Kerr-Shaw, counsel, at Skadden Arps Slate Meagher & Flom LLP.

This Practice Note on the civil liability risks of using artificial intelligence (AI) considers and summarises the following: • what is AI? • the UK Jurisdiction Taskforce (UKJT) Legal Statement on Liability for AI Harms • what are the AI civil liability risks? • who is at risk of a civil liability claim arising out of AI use? • who might bring a civil liability claim arising out of AI use? • mitigating the risk of AI generated harms and losses • AI use and the role of dispute resolution lawyers • where can further guidance and assistance be found? The matters considered in this Practice Note are limited to those arising in the England and Wales jurisdiction, albeit reference is made to developments outside the jurisdiction. For guidance on AI in the EU, see: AI (EU Law)—overview. What is AI? While the term ‘AI’ is frequently used, it can encompass a broad range of meanings. In the circumstances, when referring to AI it is

This register has been prepared in Excel and it therefore cannot be downloaded to Word. The AI risk register is illustrative and fully editable in Excel, enabling you to amend it to fit the specific needs of your organisation. Purpose of an AI risk register An AI risk register enables you to centrally record the risks to your organisation associated with using artificial intelligence (AI). You can also record how you will manage and monitor those AI risks, making it easier to keep track of any action you have taken or need to take. Where

Law360, London: Businesses that replace workers with artificial intelligence (AI) tools could face difficulties when they make claims under standard negligence insurance policies after consumers lose out, lawyers say.