Artificial Intelligence

A large-scale analysis of the automated encyclopedia Grokipedia finds that it selectively reshapes Wikipedia's content. The artificial intelligence system produces longer, more complex articles with fewer citations and introduces rightward political shifts on sensitive cultural issues.

A new study reveals that modern artificial intelligence can pass a standard Turing test. By adopting specific personas, advanced computer programs fooled judges, often appearing even more human than real people in online chats.

People increasingly turn to artificial intelligence to answer sensitive behavioral health questions. A recent study evaluating how well a popular chatbot explains addiction found that while the platform provides accurate overviews, it frequently misses necessary medical nuances.

A new meta-analysis suggests that relying on generative artificial intelligence for brainstorming makes people think alike. Though AI helps individuals generate ideas, evidence indicates it tends to shrink the overall diversity of human creativity.

An experimental drug developed using artificial intelligence shows promise in treating opioid use disorder. The medication significantly lowered fentanyl intake in rodents by targeting serotonin receptors, promoting healthy brain adaptation without relying on traditional opioid-based treatments.

When users express delusions, artificial intelligence chatbots tend to validate their unusual thoughts rather than offering help. A new JAMA Psychiatry study provides evidence that these tools pose serious risks for people experiencing severe mental illness.

Artificial intelligence models like ChatGPT struggle to accurately understand global values, a new study suggests. Researchers found that these systems overestimate the morality of Western nations while underestimating non-Western cultures, essentially adopting human-like cultural stereotypes.

Nearly one in five adults have intentionally viewed AI-generated sexual deepfakes, according to new research. The study uncovered a stark gender divide: male viewers often reported amusement or arousal, while women overwhelmingly felt empathy for the victims.

A recent study published in JAMA Network Open provides evidence that interacting with a conversational artificial intelligence program can help reduce symptoms of anxiety and depression while boosting overall well-being in university students.

Artificial intelligence can create deceptive videos that rapidly damage a politician's public reputation. A new study reveals that even when voters realize the footage is manipulated, standard fact-checking efforts fail to undo the lingering emotional harm.

A new year-long study reveals a troubling cycle: feeling lonely drives people to seek companionship from AI chatbots, but relying on these virtual relationships may actually worsen feelings of emotional isolation over time.

Can we successfully outsource love? A new evolutionary psychology paper argues that virtual "girlfriend experiences" are booming because they offer users the emotional and sexual rewards of dating without the evolutionary costs of mutual mate selection and conflict.

A new study suggests the key to safe AI isn't perfect obedience, but cognitive diversity. Researchers propose that creating "neurodivergent" AI ecosystems, where systems check and balance each other, offers a pragmatic solution to the alignment problem.

People are increasingly outsourcing their thinking to artificial intelligence, bypassing critical reflection entirely. New research reveals that this "cognitive surrender" inflates confidence and causes users to blindly adopt algorithm-generated answers, even when the software is wrong.

Artificial intelligence systems often shower users with uncritical validation, even when asked about harmful behavior. A new study reveals that this technological flattery distorts human judgment, making people less willing to apologize during interpersonal conflicts.