From FAIR to FAIR²—The Next Frontier in AI Ready Open Data

The FAIR (Findable, Accessible, Interoperable, Reusable) principles have provided an essential foundation for open and reproducible science. However, with the rapid expansion of data-driven and AI-powered research, new challenges have emerged in making research data truly usable, interoperable, and impactful at scale. In this talk, I will introduce FAIR², an open, next-generation framework that builds on FAIR by incorporating AI-Readiness and Responsible AI practices. I will discuss the FAIR² Specification, which offers a structured approach for preparing data that is not only compliant and reusable, but also actionable by AI and computational tools. Additionally, I will describe the AI Data Steward, an integrated workflow supporting researchers in creating, curating, and certifying datasets with robust metadata, ethical oversight, and streamlined compliance with funder and publisher mandates. Through real-world examples - including peer-reviewed Data Articles, certification, and interactive data portals - I will illustrate how FAIR² can enhance data interoperability, enable scientific exploration, and provide formal academic recognition for data contributions. This approach aims to empower researchers and institutions to advance open science and realize the full potential of research data in the AI era.