Lexy Kassan discusses the essential components of responsible AI, emphasizing the need for regulatory compliance and ethical implementation within the FinTech sector. She highlights the EU AI Act’s implications and the importance of aligning AI practices with ethical principles to mitigate risks while maximizing innovation.
Human-Centered AI (HCAI) emphasizes the design of AI systems that prioritize human values, well-being, and trust, acting as augmentative tools rather than replacements. This approach is crucial for ethical decision-making, bias mitigation, and fostering collaboration between humans and AI agents.
As AI becomes more powerful, protecting against its misuse is critical. This requires well-designed “safeguards” – technical and procedural interventions to prevent harmful outcomes. Research outlines a structured approach to developing and assessing these safeguards, emphasizing clear requirements, comprehensive planning, robust evidence gathering, and ongoing monitoring. This systematic process helps developers and policymakers build safer and more reliable AI systems.
The EU AI Act, introduced as the world’s first major regulatory framework for artificial intelligence, aims to create a uniform legal regime across all EU member states while ensuring citizen safety. With a risk-based approach, the Act categorizes AI systems by risk levels, ranging from low to unacceptable, and imposes strict compliance measures for high-risk applications.
On April 9, 2025, the European Commission adopted the AI Continent Action Plan, aiming to transform the EU into a global leader in AI by fostering innovation and ensuring trustworthy AI. The plan includes significant investments in AI infrastructure, the establishment of Data Labs, and the launch of strategies to accelerate AI adoption in key sectors.
The EU’s Community of Practice on Public Procurement of AI has published updated non-binding AI Model Contractual Clauses (MCC-AI) to assist public organizations in procuring AI systems. These templates are designed to align with the EU AI Act and provide guidance for both high-risk and non-high-risk AI systems.
The EU AI Act imposes strict regulations on Emotion AI, particularly in workplace settings, defining it as either “High Risk” or “Prohibited Use.” As of February 2025, the Act bans the use of AI systems to infer emotions in workplace and educational contexts, with significant penalties for non-compliance.
The EU AI Office has published the third draft of its General-Purpose AI Code of Practice, which outlines key copyright obligations for providers of general-purpose AI models. This draft emphasizes the importance of complying with EU copyright laws, particularly regarding the use of training data and the rights of content creators.
Starting from August 2, 2025, providers of General Purpose AI models will face significant obligations under the EU’s AI Act, including the need to provide technical documentation and conduct risk assessments for powerful models. The ongoing drafting of the Code of Practice raises concerns about its procedural legitimacy and the introduction of new requirements that extend beyond the original Act.
This article argues that reducing regulation for large AI providers is not a solution for Europe to catch up with the US and China, as it may only deepen dependencies on US technology. Instead, the EU should leverage its regulatory framework and investment strategy to set global standards for trustworthy AI and achieve tech sovereignty.
