The phenomenon of "AI hallucinations" – where large language models produce remarkably convincing but entirely invented information – is becoming a pressing area of study. These unexpected outputs aren't necessarily signs of a system “malfunction” per se; rather, they represent the inherent limitations of models trained on huge datasets of