• UNESCO AI ethics advisor Roy Saurabh will discuss federated learning, a technology that enables institutions to collaborate on data analysis without the need to share data directly.
• The conference will take place on April 10 at 12:00 PM in the Cloud Room of CITIC and pre-registration is not required.

The analysis of sensitive data, particularly concerning vulnerable populations such as minors, requires strict measures of privacy, security, and ethical governance. Traditional centralized data processing approaches often compromise data sovereignty and hinder compliance with regulatory standards. To address this issue, initiatives have been developed to process large volumes of information while adhering to international regulatory frameworks and ethical requirements for both quantitative and qualitative data approaches. Notable examples include UNICEF’s “Good Governance for Children’s Data” project and the architecture proposed by the EU’s EDGE project.

In this field of development, CITIC of UDC and the Integrated Engineering Group (GII) are organizing a lecture by Roy Saurabh, an expert in AI ethics and UNESCO advisor. Saurabh will discuss “Integrating Federated Learning with Data Governance Frameworks for Collaborative and Secure Sensitive Data Analysis.” The conference, held in English, will take place on Thursday, April 10, at 12:00 PM in the Cloud Room of the center.

Federated Learning: A more ethical and aecure approach to decentralized data processing

Federated learning (collaborative learning) is a type of machine learning that allows multiple institutions to work together to train AI models without sharing the original data. Instead of sending information to a single location, each institution trains the model with its own data and only shares the learning results. This helps protect privacy and keeps data at its original source while complying with security and governance standards and leveraging collective knowledge.

In this regard, Roy Saurabh’s conference examines the challenges and potential approaches for integrating federated learning with key data governance mechanisms, such as differential privacy and secure multi-party computation, to enhance data protection and regulatory compliance. By combining governance imperatives with advanced analytical methodologies, this work lays the foundation for scalable, ethical, and secure data analysis solutions.

Findings demonstrate the potential of federated learning to facilitate data-driven decision-making in sensitive environments, promoting trust, collaboration, and the ethical management of sensitive data.

Responsible AI at the Core of His Professional Career

Roy Saurabh is a senior advisor at UNESCO and a researcher in human-computer interaction (HCI), learning sciences, and AI ethics. His work focuses on transforming educational and healthcare systems through citizen science, privacy-enhancing technologies (PET), and federated learning architectures. His academic research emphasizes evidence-based interventions that uphold data sovereignty, ensure responsible AI governance, and foster human development.

Roy Saurabh is also involved in designing reliable AI assessment solutions, driving the transformation of educational systems, mental health monitoring, and digital policy-making. His contributions integrate academic research with practical implementation, influencing international governance frameworks and promoting equitable digital innovations.