Artificial Intelligence

Artificial intelligence is the broad field of building computer systems that can perform tasks normally associated with human intelligence, such as recognizing patterns in data, understanding and generating language, making predictions, classifying information, or supporting complex decisions. It is not a single technology. Machine learning, generative AI, computer vision, and natural language processing are all distinct capabilities that happen to share the AI label. That distinction matters practically: the risks, costs, data requirements, and governance needs of a generative AI writing tool are different from those of a predictive model used in credit decisions, which are different again from an image recognition system used in manufacturing. When someone says "we're using AI," the follow-up question should always be: which kind, doing what, on whose data, with what human oversight?

"AI" has become a catch-all term that can mean almost anything, which makes it easy to approve something you don't fully understand and hard to govern something you can't precisely describe. Organizations that treat AI as a single category tend to under-govern high-risk applications and over-scrutinize low-risk ones. They also tend to be surprised when a vendor's "AI feature" turns out to be a consequential decision engine running on their customers' data. The ability to ask specific questions—what capability, what data, what decision, what oversight—is not a technical skill. It is a leadership one.