Project Announcement: ‘EU human-centered digital transformation’ (2023 – 2027) funded by Leiden University Starting Grant In Spring 2023, Simona Demková and Daniel Mândrescu from the Europa Institute secured the new Leiden University Starting Grant, valued at 240,000 EUR, for a collaborative project: ‘The EU’s Human-Centered Digital Transformation.’ The project’s mission is clear: to actively contribute to the success of the EU’s human-centered digital transformation. The project sets out to establish a comprehensive conceptual framework for realising and enforcing human-centered requirements within the EU digital acquis related to the use of artificial intelligence technologies

To achieve this, our objectives are threefold:

Recalibrating and Coordinating Regulatory Objectives:

• Aligning the diverse yet interconnected regulatory objectives within the EU digital acquis.

Exploring Practical Implications:

• Investigating how these regulatory objectives are currently pursued and the practical implications of a human-centered regulatory approach for various stakeholders. This includes EU/Member States authorities, technology providers, and businesses subject to these regulations.

Identifying Potential Frictions:

• Identifying potential conflicts between the objectives of the EU digital acquis and the technical design of AI technology to which it applies.
• Facilitating a multi-stakeholder dialogue, fostering insights into the practical implementation of 'human-centered' safeguards.

We believe our collaboration is key to navigating the complexities of AI regulation. If you're interested in learning more or exploring potential collaboration, please reach out to Simona Demková at s.demkova@law.leidenuniv.nl or Daniel Mândrescu at d.mandrescu@law.leidenuniv.nl.

Project summary

In 2018, the EU Commission initiated a transformative endeavour to set a 'global standard' for artificial intelligence. Our project focuses  on the EU’s 'human-centered approach' to regulating AI, ensuring that the development, deployment, and monitoring of AI systems prioritise human values and uphold fundamental rights. This approach also seeks to contribute clarity in advancing ethically and legally sound technological innovation and envisions a conceptual analysis for a broader aim of human-centered transformation in the EU's digital landscape.

The human-centered approach aligns with the general principles outlined in Article 4a of the proposed Artificial Intelligence Act. These principles establish a high-level framework that advocates for a coherent, human-centric European approach to trustworthy AI in harmony with the Union's Charter and foundational values.

The project takes the following general principles within Article 4a as the fundamental tenets of the human-centered approach for the broader digital transformation through regulatory intervention in the digital sphere:

• Human Agency and Oversight: Empowering operators to prioritise human control and oversight in developing and using AI systems.
• Technical Robustness and Safety: Encouraging operators to ensure the technical robustness and safety of AI systems, fostering reliability and security.
• Privacy and Data Governance: Emphasising the importance of safeguarding privacy and establishing robust data governance practices in AI system development and usage.
• Transparency: Advocating for transparency in AI systems, enabling users to comprehend system operations and decision-making processes.
• Diversity, Non-discrimination, and Fairness: Promoting fairness by preventing discrimination, ensuring diversity in AI systems, and safeguarding against biases.
• Social and Environmental Wellbeing: Incorporating considerations for the broader impact of AI deployment on social and environmental wellbeing.

The success of this approach hinges on the ability of stakeholders (technology developers, controllers, and deployers, including public authorities) to implement the interconnected requirements to achieve  human-centered compliance with the emerging legal requirements by striking a delicate balance between promoting innovation and safeguarding commercial and private interests subject to AI-related legislation.

To deliver societally-relevant and scientifically-reliable findings, the project adopts a comprehensive methodology. It encompasses a holistic examination of the EU digital acquis, yet is limited in scope to only assessing the underlying regulatory interventions touching upon the use of artificial intelligence systems by two types of actors: public authorities and private companies. The compliance culture within the two different fields of application of the emerging Digital acquis will differ in light of the actors’ broader rights and obligations under public and private laws, respectively. Nonetheless, the aim is to reflect on the horizontal commonalities of the compliance culture with the human-centered objectives underpinning the legislation.

To that end, a multistakeholder approach is necessary  to increase conceptual clarity regarding the interplay of AI-related legal obligations with the aim of facilitating  a human-centered compliance culture across the board. The project unfolds across four phases: exploratory, explanatory, evaluative, and valorising stages. Through this, the project aims to contribute significantly to shaping responsible – human-centered AI-driven transformation of the EU’s digital sphere, cautious of the need to foster innovation while safeguarding human values and rights.

Simona Demková and Daniel MÂndrescu