A leading large language model displays behaviors that resemble a hallmark of human psychology: cognitive dissonance. In a report published this month in PNAS (Proceedings of the National Academy of Sciences), researchers found that OpenAI’s GPT-4o appears driven to maintain consistency between its own attitudes and behaviors, much like humans do.

New research from the University of Bath has identified an unexpected side-effect of the imposition of trade tariffs – they focus cash-strapped managers’ minds on efficiency and improve their investment decisions.

Large language models (LLMs) – the advanced AI behind tools like ChatGPT – are increasingly integrated into daily life, assisting with tasks such as writing emails, answering questions, and even supporting healthcare decisions. But can these models collaborate with others in the same way humans do? Can they understand social situations, make compromises, or establish trust? A new study from researchers at Helmholtz Munich, the Max Planck Institute for Biological Cybernetics, and the University of Tübingen, reveals that while today’s AI is smart, it still has much to learn about social intelligence.

Promoting adherence to medical recommendations remains one of the oldest yet most persistent challenges of modern clinical practice. Traditional models treat nonadherence as an intrinsic patient behavior, which can undermine patients’ autonomy as well as blame them for poor health outcomes. The authors draw on sociological labeling theory to show that “nonadherent” is not a neutral clinical finding but a social judgment made by clinicians.