Artificial Intelligence

Artificial intelligence dominates headlines, but new evidence shows most people barely use it in daily life. Researchers examined real-world web behavior and found low rates of AI use, often linked to traits like narcissism and Machiavellianism.

Can an AI chatbot combat loneliness? A study published in Disability and Rehabilitation: Assistive Technology says yes. It found that the Replika chatbot effectively reduced loneliness and social anxiety in severely lonely college students.

A new study reveals a startling preference for AI-generated medical advice over that from doctors. Participants rated AI responses as more trustworthy and complete, even when the information was inaccurate, posing a significant risk to individuals seeking health guidance.

A team of scientists has used AI to detect ADHD in adults by analyzing how their brains process rapid visual information. The findings suggest ADHD carries a unique perceptual rhythm that could help improve diagnosis and treatment.

A new study shows the type of feedback a chatbot gives affects learning and brain activity. Reflective feedback boosts problem-solving, while encouraging feedback improves memory.

Artificial intelligence chatbots are now a part of daily life, but these systems contain inherent biases from their training data. A new study investigates a pressing question: How might a chatbot's hidden political leanings influence a user's own opinions and...

A new study in Psychiatric Services finds that leading artificial intelligence chatbots respond well to suicide-related questions at the highest and lowest risk levels. However, their answers are inconsistent and vary widely when dealing with questions posing an intermediate risk.

Handing off a task to an artificial intelligence system creates a psychological distance that encourages unethical behavior. New research published in Nature demonstrates this "machine delegation" significantly lowers the moral cost for people who want to cheat for personal gain.

While many students find AI chatbots helpful for programming tasks, a recent study suggests a potential downside. The investigation discovered that students who used AI assistants more often in an object-oriented programming course tended to perform worse academically.

Professionals often view AI-assisted emails as efficient and polished, but new research finds that supervisors who overuse AI in workplace communication may be seen as lacking warmth, integrity, and competence, potentially harming trust within teams.

A new study examines how personality traits predict the use of generative AI in academic settings. The findings indicate that students who are more curious, organized, and socially outgoing tend to use AI tools more frequently for educational purposes.

Up to a quarter of brain-injured patients who appear unresponsive may actually be conscious. A new study presents an AI tool that detects this hidden awareness by tracking tiny facial movements, revealing signs of consciousness days earlier than standard clinical...

An artificial intelligence system can detect hidden signs of mild depression by analyzing facial expressions in short videos. A new study in Scientific Reports shows specific eye and mouth movements are linked to early depressive symptoms, offering a new tool...

A new study reveals that AI hate speech moderation systems are profoundly inconsistent. Different models classify identical content differently, especially for specific demographic groups, resulting in uneven and arbitrary standards for what speech is permitted online.

Artificial intelligence detected subtle movement impairments linked to Parkinson's disease by analyzing simple videos. The system identified hidden signs in early-stage patients and at-risk individuals that were completely invisible to trained neurologists, offering a new tool for early detection.