On May 29, 2026 in Paris, G7 Digital and Technology ministers approved a “Vision on AI openness opportunities and shared language”. The vision, which sets out terminology around AI openness, is the result of a three-month partnership between the OSI and the G7.
OSI Executive Director, Duane O’Brien (pictured on the monitor), presents a summary of the work done to the G7 Ministers, while highlighting the benefits of Open Source AI and the need for continued collaboration to ensure clarity.
The Vision, which comes under the French Presidency of the G7, highlights the vital role of the Open Source community in building and defining AI Openness. Additionally, it recognizes the challenges in understanding Openness and Open Source in the context of AI, as well as the elements that determine openness, and it calls for the use of clear and appropriate labels that accurately describe the degree of openness of AI systems.
To this end, the Vision sets out clear criteria for AI model openness: labeling models with proprietary licensing as “Weights Available”, while labeling models which are distributed under an Open Source license as “Open Weights”. When it comes to Open Source AI, the criteria broadly follow those of the OSI’s Open Source AI definition (OSAID), however restrict exceptions to publishing training data to cases of legal or technical impossibility, and only require data information in absence of training data. Finally, the Vision adds criteria for “Open Source AI with Open Data” to describe AI systems where all assets are released free of charge under an Open Source license – including its models’ weights, deployment code, training code, and full training data.
