EU

EU's Community of Practice on AI has released updated non-binding Model Contractual Clauses (MCC-AI) for public procurement of AI systems. Two templates address “high-risk” and “non-high-risk” AI systems, aligned with the EU AI Act. This guidance aims to assist public organizations but may also benefit private companies. Stakeholders are encouraged to report their use of MCC-AI, enhancing AI procurement practices in the public sector.

https://www.insideprivacy.com/artificial-intelligence/eus-community-of-practice-publishes-updated-ai-model-contractual-clauses/

EU AI Office's third draft of the General-Purpose AI Code of Practice outlines commitments for GPAI model providers regarding copyright compliance under the AI Act, effective August 2025. Key obligations include adhering to training data copyright laws, respecting opt-out requests from content creators, and limiting web crawling practices. The streamlined draft emphasizes transparency and governance measures, with a focus on mitigating copyright infringement risks. Differences in US and EU copyright practices, such as the lack of a “fair use” doctrine, are noted, highlighting the complexities of navigating AI copyright law in Europe. Finalization expected May 2025.

https://www.morganlewis.com/pubs/2025/04/eu-ai-office-publishes-third-draft-of-eu-ai-act-related-general-purpose-ai-code-of-practice-key-copyright-issues

EU's push for AI competitiveness led to withdrawal of AI Liability Directive (AILD), raising concerns about accountability in AI-related harms. Big Tech benefits from this retreat, avoiding liability for potential damages. Effective oversight becomes challenging due to AI's ‘black-box' nature, risking consumer protection. A call for reassessment of AI regulation instead of deregulation is essential for safeguarding citizens against harmful practices.

https://euobserver.com/digital/arcbd1284c

EU's AI Office launched a survey to collect AI literacy practices for a living repository, currently featuring over 20 examples. This initiative, aiming to enhance AI literacy and support the AI Act's Article 4, invites organizations to share experiences. Contributions will be verified for transparency before inclusion. The repository serves to foster collaboration and learning among AI providers.

https://digital-strategy.ec.europa.eu/en/news/new-survey-gather-practices-ai-literacy-living-repository

EU AI law faces backlash: Industry pressures lead to voluntary systemic risk assessments in the Code of Best Practices. Reporters Sans Frontières exits talks, citing industry influence and lack of protections for information rights. EU Parliament members express concerns over weakened regulations, risking fundamental rights and democracy.

https://www.eunews.it/en/2025/04/02/key-part-of-eu-ai-law-under-attack-from-hi-tech-industry/

RSF withdraws from European AI Act negotiations due to lack of protections for the right to information and influence from tech giants, condemning the code as insufficient and sidelining democratic issues. They stress the need for concrete provisions to combat misinformation and safeguard reliable information.

https://rsf.org/en/rsf-withdraws-negotiations-european-ai-act-s-code-practice

EU AI Act compliance begins February 2025, with early requirements like AI literacy and definitions of AI systems effective immediately. Organizations must educate personnel on AI risks and integrate literacy into governance frameworks. Key prohibitions include exploiting vulnerabilities and social scoring. Businesses should assess AI systems, ensure training, and stay updated on regulatory changes for compliance. Proactive approaches in governance may enhance trust and competitive positioning in the AI landscape.

https://kennedyslaw.com/en/thought-leadership/article/2025/first-steps-to-compliance-meeting-early-obligations-under-the-eu-ai-act/

EU AI Act needs risk-based approach: The EU should avoid a one-size-fits-all AI regulation, which risks stifling innovation due to overlapping enforcement and lack of risk/benefit analysis. A sector-specific approach is suggested to balance risks and benefits, streamline regulation, and enhance competitiveness, ensuring that future AI guidelines support innovation while safeguarding public interest.

https://project-disco.org/european-union/balancing-ai-innovation-and-regulation-a-risk-based-approach/