White paper edited by Christine Legner, Diego Kuonen, and André Golliez: “Artificial Intelligence (AI) is rapidly transforming how organisations across business, research, education and the public sector operate, innovate and create value. Yet, the effective and trustworthy use of AI does not depend on algorithms, models and computing infrastructure alone. AI cannot operate without data, and AI-ready data are an important foundation for trustworthy AI systems. They help reduce risks associated with unreliable, biased or inappropriate data and improve the quality and reliability of AI outputs. Documented data quality, provenance, meaning and permitted uses also help building trust in AI systems among users and other stakeholders.
The importance of data is also reflected in the practical challenges of AI projects: more than 80% of the time spent on an AI project may be devoted to data, while 72% of organisations report data management as a key challenge in scaling AI. Moreover, even among high-performing organisations, 70% report difficulties in developing processes for data governance and integrating data into AI models quickly. The availability of data that are fit for their intended use is therefore fundamental for developing, operating and improving AI systems and for helping build trust in their use. Against this background, the AI Action Plan for Switzerland4 brings together concrete actions aimed at strengthening Switzerland’s readiness for AI across five key areas. One of these is AI-ready Data, with the overarching objective of “ensuring the availability of high-quality, appropriately documented and augmented datasets for training or running AI systems and strengthening their provision and use across business, research, education, and the public sector”…(More)”.
