Open source technology in the age of AI
McKinseyWith more organizations deploying gen AI across business functions, a new survey finds that leaders are increasingly turning to open source AI solutions to build out their tech stacks.
With more organizations deploying gen AI across business functions, a new survey finds that leaders are increasingly turning to open source AI solutions to build out their tech stacks.
However, the company has called the Llama 4 models open weight and not open source, after receiving criticism that it does not match the Open Source Initiative’s official definition. This is mainly because Meta does not reveal the data that the AI models were trained on.
They’re available for download “In keeping with our commitment to open source” according to Meta’s announcement post. However the Open Source Initiative has claimed the Llama 4 Community Model is not open source because users in the European Union are denied some rights offered to users elsewhere.
Many firms are misusing the ‘open source’ label. The scientific community, which relies on transparency and replicability, must resist this trend.
“The current landscape of AI model licensing is riddled with confusion, restrictive terms, and misleading claims of openness,” Vidal said. “Instead of redefining ‘open’ to suit corporate interests, the AI industry should align with established open source principles to create a truly open ecosystem.”
The United Nations Open Source United community and the Open Source Initiative (OSI)Opens a new window today announced that the OSI has become the first organization to officially endorse the UN Open Source Principles. The UN Open Source Principles, recently adopted by the UN Chief Executive Board’s Digital Technology Network (DTN)Opens a new window, provide guidelines to drive collaboration and Open Source adoption within the UN and globally.
It’s currently unclear whether DeepSeek’s planned open source release will also include the code the team used when training the model. That kind of training code is necessary to meet the Open Source Initiative’s formal definition of “Open Source AI,” which was finalized last year after years of study. A truly open AI also must include “sufficiently detailed information about the data used to train the system so that a skilled person can build a substantially equivalent system,” according to OSI.
In traditional software, the perennial struggle between open source and proprietary revolves around the “true” meaning of “open source.” This can be resolved by deferring to the formal “definition” as per the Open Source Initiative, the industry stewards of what are and aren’t legitimate open source licenses. More recently, the OSI has formed a definition of “open source AI”
“The point of having definitions is to have criteria that can be scored, and focusing on licensing is how that is accomplished,” Maffulli said in a statement issued to TechCrunch. “The global community and industry have come to rely on the Open Source Definition and now the Open Source AI Definition as objective measures that they can rely on.”
OSI executive director Stefano Maffulli told me. “If only Meta’s license would remove the restrictions, we’d be more in sync. As it stands now, Llama is a liability for any developer; too opaque to be safe to use and with a license that ultimately leaves Meta in charge of their innovations.
Fontana also warns against overreach in defining openness, advocating for minimal standards rather than utopian ideals. “The Open Source Definition (OSD) worked because it set a floor, not a ceiling. AI definitions should focus on licensing clarity first, not burden developers with impractical transparency mandates.”
For an AI model specifically to be considered truly open, the OSI says its developers must provide detailed information about its training data and give people the ability to study, use and modify the system for any purpose.
