The article investigates how generative AI technologies threaten to reshape, and potentially erode, the cognitive foundations of critical thinking. While digital tools increasingly mediate learning, reflection, and communication, their capacity to generate fluent text risks displacing essential practices—such as argumentative writing and handwriting—that are foundational for the development of analytical and independent reasoning. Drawing on cognitive science, philosophy of education, and recent empirical studies, we argue that generative AI facilitates a form of cognitive outsourcing that undermines epistemic engagement and reflective effort. The article offers a dual-level analysis: on one hand, it shows how writing practices constitute thought rather than merely express it; on the other, it explores how LLMs may bypass the deliberative and justificatory processes essential to critical autonomy. We also review emerging pedagogical strategies aimed at integrating AI tools without compromising cognitive growth, and we conclude with a normative call to protect the conditions—both technological and educational—under which critical thinking can survive. The article thus contributes to an urgent philosophical debate: whether the widespread adoption of generative AI is compatible with the preservation of intellectual self-determination in democratic societies.
