Democratizing AI: Leveraging Open Data and FOSS Tools to Build Transparent Machine Learning Systems

Machine Learning has seen rapid adoption in recent years, but much of it remains locked behind proprietary systems, commercial datasets, and opaque algorithms. This talk aims to highlight how Free and Open Source Software (FOSS) and open data are transforming that narrative—making AI more transparent, accessible, and ethical.

In this 15-minute lightning talk, I will walk through how anyone can build meaningful ML models using only open datasets and widely available open-source tools. I’ll demonstrate this with a simple yet impactful project, such as predicting air quality using open government environmental datasets. The project utilizes popular FOSS libraries like Scikit-learn or PyTorch, with all code and data made publicly available.

Key points will include:

• The importance of open datasets in democratizing AI

• How FOSS tools enable reproducible and ethical ML workflows

• Tips for new contributors to get involved in open AI projects

• The role of the community in auditing, improving, and documenting models

By the end of the session, attendees—especially students and early-career developers—will leave with an understanding of how they can start building open ML solutions and contribute to the FOSS AI ecosystem, regardless of their background or institutional access.

• Understand how open datasets and FOSS tools make machine learning accessible to everyone.

• Learn how to build a simple ML project using only open-source libraries and public data.

• Gain awareness of ethical and transparent AI practices through community-driven development.

• Discover ways to contribute to the FOSS AI ecosystem—beyond coding, including data curation, documentation, and sharing models.

• Get inspired to start your own AI projects using freely available resources and contribute back to the community.