Open Source Initiative Blog

1. Why datasets built on public domain might not be enough for AI

   Common Corpus is a public domain dataset for training large language models (LLMs). Boasting 500 billion words in multiple languages, drawn from various cultural initiatives, it offers researchers a powerful tool to develop smaller and more efficient LLMs. It should not be abused as a tool to promote public policies that expand the reach of copyright law.

2. Open Source AI Definition – Weekly update May 6

   With a call for application, a town hall meeting and a lot of comments on the 0.0.8 definition, this past week was busy. Catch up here!

3. CRA standards request draft published

   The European Commission recently published a public draft of the standards request associated with the Cyber Resilience Act (CRA). For those who depend on incorporating or creating Open Source software, there is an encouraging new development found here. For the first time in a European standards request, there is an express requirement to respect the needs of Open Source developers and users.

4. Open Source AI Definition – Weekly update April 29

   With a new 0.0.8 draft definition, discussion on a new OSI license and and FAQ page, last week was busy! Get your update here.

5. Openly Shared: CRA’s Open goes beyond the OSD

   The definition of “open source” in the most recent version (article 2(48)) of the Cyber Resilience Act (CRA) goes beyond the Open Source Definition (OSD) managed by OSI.