Mozilla releases report on open-source AI landscape

Mozilla has published a report on open-source AI, providing primary data and analysis for engineers building AI tooling [Mozilla Report].

The report lists the most widely used open-source frameworks—TensorFlow, PyTorch, and Hugging Face Transformers—and the top libraries such as NumPy, Pandas and scikit‑learn. It quantifies developer challenges: 42 % of respondents cite insufficient documentation, while 37 % point to limited community support [Mozilla Report].

Key findings show that transparency and explainability are increasingly emphasized, with 58 % of projects allocating dedicated effort to these areas. Dataset diversity remains a concern; 31 % of projects report a need for more representative data. Adoption metrics reveal TensorFlow in 68 % of surveyed projects and PyTorch in 55 % [Mozilla Report].

The data underscores that open-source AI continues to expand, yet the report flags sustainability questions as proprietary solutions gain market share. Developers and organizations looking to adopt or contribute to open-source AI can use these metrics to benchmark their own efforts and prioritize documentation, community engagement, and data diversity.

Poll: Which open-source AI framework do you prefer?

• TensorFlow
• PyTorch
• Scikit-learn