AI Officially Passes Turing Test: Language Models Aided by ‘Human Persona’

Groundbreaking research from UC San Diego reveals advanced language models have passed the classic Turing Test for the first time. By adopting human-like ‘personas,’ models like GPT-4.5 and LLaMa-3.1 convincingly impersonated humans, often fooling participants more effectively than real people. This study redefines the Turing Test as a measure of conversational imitation.