New study views large language models as 'cognitive viruses' and warns of cognitive dependency

On September 9, 2026, IT Home reported that an international research team published a paper suggesting that the spread of AI mirrors the spread of viruses. The paper, not yet peer-reviewed, proposes a new theoretical framework to explain why people are increasingly reliant on AI tools and how this dependency is reshaping societal cognition. The study likens large language models to 'cognitive viruses,' arguing that these technologies spread through cultural transmission due to their utility, yet risk outsourcing human thinking. The authors warn that this dependency could weaken cognitive abilities, particularly among students who rely on AI for critical thinking. The paper emphasizes the need for 'cognitive immunity' through practices like independent problem-solving and critical discussion. Researchers stress that immunity does not mean avoiding AI, but rather fostering cognitive engagement and maintaining skills that do not rely on external tools.