European AI Trend Radar: Summary & Overview

A critical need for tailored educational programs was identified to bridge the gap between technical developers and end-users, particularly in sectors like healthcare, manufacturing, and public administration. For example, AI education in healthcare aims to help professionals understand AI’s potential and limitations, thereby fostering trust and improving patient outcomes.

The workshops emphasized the importance of effective communication strategies to build public trust in AI technologies. For instance, informing the public about AI’s role in healthcare through transparent communication can lead to better acceptance and utilization of AI-driven solutions.

Interdisciplinary research and collaboration, such as between AI experts and public sector workers, are essential for integrating AI into various sectors. This was seen in the energy sector, where AI competency courses were recommended to address gaps in education and promote knowledge transfer.

A major challenge highlighted was the need to overcome information silos across organizations to enable better data sharing. For example, in healthcare, a European-wide data infrastructure could advance research by providing access to large, diverse datasets.

The use of synthetic data was discussed as a solution for environments where real-world data is scarce, such as in manufacturing where defects are rare but critical for training AI models.

Trustworthy AI was identified as crucial, encompassing aspects like robustness, security, explainability, and ethics. For example, AI systems in manufacturing must be robust to changing work conditions, ensuring reliable performance.

Ensuring AI systems are safe and reliable, particularly in critical environments like space missions, requires advanced verification techniques. This is critical in high-stakes industries where failure can have catastrophic consequences.

The need for systematic approaches to measure the impact of AI on society was emphasized, including evaluating AI’s societal and economic effects. For instance, in manufacturing, understanding how AI affects productivity can guide ethical and effective AI integration.

The TDWs highlighted the importance of joint labs and collaboration between academia and industry. For example, in manufacturing, joint labs can accelerate the practical application of AI, bridging the gap between research and real-world industrial needs.

Governments play a significant role in shaping AI markets through procurement and market creation. Strategic decisions between in-house development and private procurement can stimulate innovation. The TDWs discussed how public-private partnerships and updated procurement guidelines can enhance AI adoption, especially in sectors like healthcare.

The creation of algorithm registers was identified as a way to enhance transparency and accountability in AI systems. This would involve documenting AI algorithms used in systems, making them visible and open to scrutiny. This is crucial for building trust, particularly in sectors where AI decisions have significant consequences.

Federated Learning (FL) was highlighted as a promising approach in manufacturing, allowing the training of AI models on partial datasets across multiple locations without sharing raw data. This method aligns with privacy regulations like GDPR and is effective in sectors handling sensitive information.