Study: Identical Behavior Doesn't Make AI Seem Conscious — SkimNews
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- Ludwig-Maximilians-Universität München (LMU) researchers published a Cognition study finding that participants consistently rated AI as less conscious than humans even when both behaved identically — the most responsive AI scored below the least responsive human on consciousness.
- The research, led by Dr. Louis Longin alongside Dr. Bahador Bahrami and Professor Ophelia Deroy, surveyed nearly 1,100 participants who evaluated characters explicitly labeled as either human or AI across matched behavioral scenarios.
- When participants were asked about "awareness" rather than "consciousness," their judgments of humans and AI followed largely similar patterns, suggesting the two terms trigger fundamentally different mental categories despite overlapping everyday use.
- The team attributed the gap to essentialism: people treat consciousness as tied to subjective inner experience, while awareness is read as a neutral descriptor of an agent detecting its environment.
- Senior author Bahador Bahrami warned that companies and journalists routinely "reach for hyped descriptions" crediting AI with intentions, plans, or moral conscience — language that, the data suggests, outpaces how everyday people actually judge these systems.
Why it matters: The study provides the first direct head-to-head test of how identical behavior is judged differently based on whether the actor is labeled human or AI. Per the authors, this matters because public-facing AI discourse leans heavily on consciousness-laden language — words the data shows people don't actually apply to machines even when behavior matches, with the human-AI gap narrowing or disappearing only under more neutral terms like "awareness."
Ask SkimNews
