Something doesn't need to be full human-level general intelligence to be considered as falling under the "AI" rubric. In the past people spoke of "weak AI" versus "strong AI" and/or "narrow AI" vs "wide AI" to reflect the different "levels" of AI. These days the distinction that most people use is "AI" vs "AGI" which you could (very loosely) speaking think of as somewhat analogous to "weak and/or narrow AI" vs "strong, wide AI".