GLOSSARY
Open source AI
Under the Open Source Initiative's definition, open-source AI gives broad freedoms to use, study, modify and share a system. It requires suitable access to code, model parameters and information about training data. An openness claim should be checked against a named definition and the actual terms.
What procurement and sales need to establish
A company may want to operate a model itself or build an industry application around it. Permitted use, redistribution and modification then matter directly. A general promise of openness does not establish whether that deployment is allowed or technically manageable.
Openness can improve switching options and make components easier to inspect. It also creates responsibilities for operation, updates and security assessment. Someone must own those responsibilities. An inexpensive or freely available model does not remove the need to examine total cost of ownership.
Verify the definition and the deployment
Selection should cover the licence, available components and documented restrictions. A practical exercise can then show whether the team can make its intended change and sustain operation. Accessible model weights alone do not answer every question about rights and supporting components. A specific licensing assessment still depends on the intended use.