Ludwig Maximilian University of Munich Researchers Investigate AI Consciousness
A Ludwig Maximilian University of Munich study published in Cognition reveals people do not view artificial intelligence as conscious even when it exhibits human-like behavior. Led by Dr.
Whereas previous studies have typically asked general questions about whether AI actually has mental states, our study is the first to directly compare how people attribute the same mental states to AI and humans behaving in exactly the same way, under identical circumstances,
says Longin, lead author of the study.
Experimental Design and Methodology Behind the LMU Study
The study tested nearly 1,100 participants through distinct experimental groups. One cohort read brief scenarios detailing artificial agents displaying varying degrees of environmental responsiveness, such as reacting to ambient sounds or recognizing emotional cues in others. A second cohort read identical scenarios featuring human protagonists instead of AI systems. Following these readings, participants rated the degree of consciousness and awareness each character seemed to be of its surroundings.
Responsiveness directly shaped perceptions of both human and machine subjects. Highly reactive agents earned higher scores for general awareness and consciousness alike. Yet a distinct gap emerged when evaluating consciousness in strict isolation. Even maximally responsive artificial intelligence models rated lower on the consciousness scale than the least responsive human subjects tested.
Differentiating Awareness From Subjective Experience
The experimental outcomes shifted when researchers substituted the term consciousness with awareness. Participant ratings for awareness tracked along remarkably similar trajectories for both human and machine subjects, suggesting a fundamental divergence in how individuals conceptualize these two mental states.
People are quite willing to say that an AI notices things and is aware of its surroundings. But they clearly draw a line when it comes to the term ‘consciousness’. People consistently attribute less consciousness to AI than humans — even when their behavior is identical.
As Dr. Louis Longin noted, public discourse often worries that human-like machine behaviors will trick users into ascribing genuine mental states to software. However, the data indicates a much more nuanced cognitive boundary.
Implications for Public Discourse and Tech Communication
The findings offer concrete direction for how researchers, journalists, and enterprise developers discuss advanced software models. Professor Ophelia Deroy pointed out that mental terminology often creeps into tech descriptions simply because language lacks superior alternatives for models trained to mirror human speech and action.
We may be using mental terms to refer to AI, but this may only be because we lack better words. After all, AI is trained to act and speak like a human, so these descriptions seem natural,
says Ophelia Deroy, philosopher and a senior author of the study.
Dr. Bahador Bahrami emphasized that everyday public judgments display high discrimination, resisting simple equivalence between human cognition and algorithmic execution. Because neutral descriptors like awareness bridge the gap between human and machine evaluations far more fluidly than loaded psychological terms, the vocabulary chosen by developers and media outlets carries substantial weight in shaping public understanding.
Companies and journalists often reach for hyped descriptions of AI models, and credit them with intentions, plans, or even moral conscience and hesitation. What our study shows is that everyday judgements are much more discriminating: People do not simply put AI and humans on the same mental scale. But the boundary also depends on which terms we use,
