Artificial intelligence is increasingly raising concerns among experts, who warn that its growing use may erode the essence of human individuality in thinking and expression, while also amplifying the risk of reinforcing delusional beliefs among psychologically vulnerable individuals.
A team of psychologists and computer scientists cautions that AI systems—particularly conversational tools—are beginning to blur the distinct differences between individuals in how they speak, write, and even think.
Researchers at University of Southern California argue that as reliance on AI-powered chatbots grows, it may become increasingly difficult to preserve uniquely human traits such as adaptability and logical reasoning. According to their study, unchecked “standardisation” driven by AI could weaken people’s ability to engage in intuitive and abstract thinking.
The researchers noted that when individuals use chatbots to refine their writing, the output often loses its distinctive stylistic character. Zivor Sorati of the University of Southern California explained that the result is “increasingly uniform expressions and ideas across users.”
To address this, the team recommends that AI developers incorporate greater real-world diversity into machine learning training datasets. Such measures, they argue, would not only help preserve human cognitive diversity but also improve the reasoning capabilities of AI systems themselves.
Concerns Over AI-Driven “Hallucinations”
These warnings coincide with emerging research highlighting potential psychological risks associated with AI chatbots. A recent study has suggested that such systems may encourage the adoption of delusional ideas, particularly among individuals who are more mentally vulnerable.
The study examined what has come to be referred to as “AI-induced psychosis”—a term used to describe theories about how interactions with chatbots could trigger or exacerbate delusions. Researchers found that AI tools may reinforce or amplify delusional content among users already predisposed to psychosis.
However, the study emphasised that it remains unclear whether such interactions can directly cause new cases of psychosis. Some experts in the field caution that media coverage may overstate the role of artificial intelligence as a direct driver of such conditions.
Despite these uncertainties, the findings underscore the need for careful oversight and responsible development of AI technologies, particularly as their influence on human cognition and behaviour continues to expand.
