AI disclosure labels may do more harm than good, study warns

The growing use of AI-generated science and science-related content, particularly on social media, raises important concerns: such texts may contain false or highly convincing information that is difficult for users to detect and could influence public opinion and decision-making.