General-purpose AI model
Also known as: Foundation models, Base models, Large Language Models (LLMs)
A general-purpose AI model is an artificial intelligence system trained on broad data sets that can be adapted to perform a wide variety of different tasks. Unlike narrow AI, which is built for one specific function, these models provide a versatile base that can be fine-tuned or prompted for numerous applications across various domains.
In practice
Auditors encounter these when reviewing "shadow AI" usage where employees use public LLMs to process sensitive company data without authorization. They also appear during third-party risk assessments of software vendors who integrate foundation models as the core engine of their product.