The Pinocchio Effect: Why AI Lies Without Knowing

This essay introduces the concept of the Pinocchio Effect in generative artificial intelligence: the production of elegant, convincing, but false content without the intention to deceive. Through metaphors, real-world examples, and academic references, the paper explores the anatomy of AI hallucinations, their architectural roots in large language models, and the cognitive and social risks of persuasive falsehoods. Finally, it proposes strategies of epistemic hygiene and critical thinking to resist the seduction of well-crafted lies, highlighting the responsibility of humans in preserving truth in the age of AI.