The paper explores the impact of the *European Union Artificial Intelligence Act* (AI Act) on Federated Learning (FL), a privacy-preserving machine learning approach. The AI Act, which aims to promote trustworthy AI while protecting fundamental rights, poses significant challenges and opportunities for FL. The authors conduct an interdisciplinary analysis of the AI Act's legal and ML implications, focusing on data governance, energy efficiency, robustness, and quality management. They find that FL can address many of these challenges, particularly in terms of data privacy and access to siloed data. However, FL also faces new performance and energy efficiency issues, such as the need for frequent validation and the trade-off between privacy and energy consumption. The paper suggests that FL could become a crucial component of AI Act-compliant ML systems if the community shifts its research priorities to align with the Act's requirements. Key recommendations include improving data quality techniques, optimizing energy efficiency, and developing a technical framework for regulatory compliance. The authors advocate for a redirection of FL research to ensure it meets the societal values encapsulated in the AI Act.The paper explores the impact of the *European Union Artificial Intelligence Act* (AI Act) on Federated Learning (FL), a privacy-preserving machine learning approach. The AI Act, which aims to promote trustworthy AI while protecting fundamental rights, poses significant challenges and opportunities for FL. The authors conduct an interdisciplinary analysis of the AI Act's legal and ML implications, focusing on data governance, energy efficiency, robustness, and quality management. They find that FL can address many of these challenges, particularly in terms of data privacy and access to siloed data. However, FL also faces new performance and energy efficiency issues, such as the need for frequent validation and the trade-off between privacy and energy consumption. The paper suggests that FL could become a crucial component of AI Act-compliant ML systems if the community shifts its research priorities to align with the Act's requirements. Key recommendations include improving data quality techniques, optimizing energy efficiency, and developing a technical framework for regulatory compliance. The authors advocate for a redirection of FL research to ensure it meets the societal values encapsulated in the AI Act.