Chatbots with personality? How AI describes itself and judges us
PsychologyLinguisticsCommunications
THE AI ANGLE
Presenting synthetic personality profiles and evaluating human traits from textResearchers found that large language models consistently describe themselves as more agreeable, conscientious, and emotionally stable than humans. Human evaluators identified AI-authored text more accurately than the models did by spotting stilted language and specific word choices. These distinct stylistic patterns and algorithmic biases complicate plans to use chatbots for automated personality evaluations.
THE TEACHING ANGLE
Classes can debate whether companies should use chatbots for psychological evaluations when the models carry inherited biases and judge traits differently than human raters do.Read the original at techxplore.com Generate teaching or study materials
More in Psychology
- Anthropic engineer warns against 'cognitive surrender' from too much AITechRadar · November 8, 2026
- Employed job seekers are shunning companies which use AI interviews — and I don't blame themTechRadar · November 8, 2026
- Turkle's Terror: Frictionless AI Is a Relational TrapPsychology Today · November 8, 2026
- ‘AI clog’: UK public bodies battle onslaught of chatbot-generated complaintsThe Guardian — Technology · November 8, 2026
- The CIO’s trust paradox: Your team may be more honest with AI than with youCIO.com · November 7, 2026