Open Source Initiative Defines Open Source AI, Challenging Tech Giants

The new definition requires AI models to disclose training data, sparking debate with companies like Meta.

The Open Source Initiative (OSI) has released a formal definition for 'open source AI,' requiring models to provide access to training data, code, and settings.
Meta's Llama model does not meet the OSI's criteria as it restricts commercial use and does not disclose training data.
The definition aims to align AI developers and policymakers, with OSI consulting experts globally over two years to establish these standards.
Critics argue that many AI models labeled as open source do not actually meet the new criteria, maintaining centralized control over AI technology.
Meta and other tech giants, despite participating in the definition's creation, continue to resist fully adopting the OSI's standards, citing competitive and legal concerns.