Closed Source
Closed-source software does not provide its source code under terms that grant open-source rights to study, modify and redistribute it. In artificial intelligence (AI), code, model weights and training materials can each have different access and licensing terms.
Suppose a provider offers an AI service through an application programming interface (API) but does not release its implementation. A customer can submit requests under the service terms without gaining the source code needed to modify or reproduce the service. That is an example of a closed implementation, not a description of every hosted API.
Hosting and licensing are separate. An open-source application can be offered as a paid hosted service. A downloadable model can also have restrictive terms, and publicly visible code can lack permission to redistribute modifications. Check each artifact and its license rather than inferring openness from price or access method.
This matters for inspection and operational control. If a team cannot run or modify an implementation, it depends on what the provider exposes through documentation and interfaces. Open weights grant access to parameters, subject to their terms; that does not automatically establish open training code or an open-source AI system.
Sources
- Open Source Initiative: The Open Source Definition — States the source and licensing rights that a release must provide to meet the software definition.
Go deeper
- Open Source Initiative: Approved licenses docs
Compare concrete license permissions instead of inferring rights from visible code.