Introducing the FAIR² Specification

FAIR
AI-Ready
Responsible AI Aligned
Context-Rich
FAIR² (FAIR Squared™) extends the FAIR principles with a formal specification that makes research data AI-ready, responsibly governed, and optimized for deep scientific reuse. It provides a context-rich representation of each dataset, ensuring rigor, reproducibility, and interoperability. Compatible with MLCommons Croissant, FAIR² integrates with TensorFlow, JAX, and PyTorch, enabling AI-driven analysis and broad data sharing on platforms like Kaggle and Hugging Face.


An initial release of the FAIR² Specification will be available soon.
Learn more About FAIR²

What is FAIR²?

FAIR² (FAIR SQUARED) builds on the FAIR principles to enhance scientific datasets with context-rich metadata, Responsible AI alignment, and AI-ready compatibility.
Context-Rich Metadata: Fully document how data is created, processed, and validated.
AI-Ready Design: Ensure compatibility with modern workflows and machine learning.
Responsible AI Alignment: Transparently address biases, ethics, and limitations.

Why Choose FAIR²?

FAIR² provides essential advancements for modern science
Discoverability with Depth: FAIR² datasets are not only easy to find but also easy to understand, thanks to detailed metadata and provenance.
Trust Through Transparency: By documenting workflows, ethical reviews, and biases, FAIR² builds confidence in data quality and usability.
Future-Proof Standards: FAIR² adapts data practices for the growing demands of interdisciplinary collaboration and AI integration.

Development Roadmap

FAIR² is designed with the scientific community, evolving to meet shared goals for transparency, reproducibility, and ethical alignment.
Current Progress: An early preview of the FAIR² Specification  will be available soon for review. Early discussions on establishing community-based governance are underway.
Next Steps: Establishing community governance in 2025 and preparing for formal FAIR² Certification processes.
Mailing Address
FAIR² @ SENSCIENCE

Avenue Tribunal-Federal 34
1005 Lausanne,
Switzerland
Terms & Conditions
Pages
About
Specification
Get Involved
Contact
Contact
info@fair2.ai