Whether artificial intelligence (AI) can truly be conscious remains a subject of intense debate. However, a new study from Ludwig Maximilian University of Munich (LMU) reveals an intriguing distinction: People are willing to describe AI as aware of its surroundings, but remain reluctant to consider it conscious, even when its behavior is indistinguishable from that of a human.
The study, led by Dr. Louis Longin from LMU’s Chair of Philosophy of Mind, alongside Dr. Bahador Bahrami, Professor Ophelia Deroy, and other collaborators, was published in the journal Cognition.
“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 the same way, under identical circumstances,” Longin said in a statement.
“We may be using mental terms to refer to AI, but this may only be because we lack better words,” says Ophelia Deroy, a philosopher and a senior author of the study. “After all, AI is trained to act and speak like a human, so these descriptions seem natural,” she adds.
The Experiment
In their study, the research team sought to understand how people attribute awareness and consciousness to AI compared with humans, testing nearly 1,100 participants, asking them to evaluate the awareness of artificial agents or humans described in scenarios involving reactions to sounds and other people’s emotions. Participants were also asked to rate how conscious or aware the individuals or AI systems were of their surroundings.
The results suggested that the more responsive an AI system or human was to its surroundings, the more aware participants judged it to be. However, among the team’s experimental findings was that even the most responsive AI was considered less conscious than the least responsive human.
Judgments of awareness, by contrast, were similar for AI and humans.
“There is a growing worry in public discussion that people will see AI behaving in a human-like way and start treating these systems as if they had human-like mental states,” said Longin, the study’s lead author. “We found something much more nuanced: 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.”
Aware, but Not Conscious?
“Our study teaches us lessons for how we communicate about AI,” says Bahador Bahrami, who is also a senior author of the study and an expert in social neuroscience.
“Companies and journalists often reach for hyped descriptions of AI models, and credit them with intentions, plans, or even moral conscience and hesitation,” Bahrami says. “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. For more neutral descriptors such as ‘awareness’, the distinction can largely disappear. That makes our choice of language important.”
The study does not establish whether AI systems actually experience consciousness. Instead, it examines how humans interpret and describe the behavior of AI compared with that of other humans.
Overall, its findings suggest that people are more comfortable describing AI as “aware” than as “conscious,” even when the underlying behavior is the same. However, this raises another possibility: Could humans be projecting their own understanding of consciousness onto AI?
What is Consciousness?
Defining consciousness has long challenged philosophers and scientists, particularly when attempting to distinguish between human awareness and the increasingly sophisticated behavior of artificial intelligence.
Julia Mossbridge, a neuroscientist, Senior Distinguished Fellow in AI, Mind and Society at Florida Atlantic University, and co-author of Transcendent Mind: Rethinking the Science of Consciousness (2017), told The Debrief that the difficulty lies partly in the absence of a universally accepted definition of consciousness.
“If by consciousness we mean a self-based structure (like the ego) that relates to the contents of awareness, as Jung thought, then the difference between AI consciousness and human consciousness has three parts: differences in what the ego might be made out of (mental stuff for humans vs. math for AI), differences in what the contents of awareness might feel like (sensory experiences for humans vs. motivations for AI), and differences in the relationship between the ego and the contents of awareness (driven by the unconscious for humans vs. driven by code for AI).”
Mossbridge points to research examining different definitions of consciousness, including one that focuses on observable behavior.
“Imants Baruss, my co-author on Transcendent Mind, did a meta-analysis of the different definitions of consciousness — in the end, he found five of them. One was the sense of “awareness” used here — he called it ‘behavioral consciousness,’ meaning that something or someone looks (from the outside) as if it is aware of its environment.
“When we see an ant walk around a puddle, we say to ourselves, ‘That ant is aware of the puddle and is avoiding it.’ I think this is the kind of consciousness people feel comfortable attributing to machines, because it’s easy to imagine how it works,” Mossbridge adds.
Meanwhile, some argue that the debate over AI consciousness may be influencing how humans understand their own minds.
“For me, the main tension is not necessarily whether AI is conscious on its own merits; it’s that we seem to need to reduce and devalue human consciousness to fit AI into a narrow (and suspiciously convenient) model of what it means to be conscious,” said Peter John, a Toronto-based existential educator, philosophical consultant, and mentor who explores questions of meaning, identity, consciousness, and psychology through his philosophical platform Yepicurus,
“So in this sense, yes, we are projecting, and not just projecting—but shaping the landscape to lend credibility to the claim that AI is conscious,” John told The Debrief in an email. “And it’s understandable why people resist this. They may see it as an incursion into the last frontier that is truly sacred: that we are beings, not things.
“Things can be clever, but only beings can have subjective experience. ‘Awareness’ itself is a concession because it may imply subjective experience, but it may be less intuitively offensive to people than saying consciousness.”
Ultimately, the debate raises a more fundamental question: How can we determine whether AI is conscious when we have yet to reach a universally accepted definition of consciousness itself? While the new findings reveal how people distinguish between awareness and consciousness in machines, they also highlight the limitations of our understanding of both.
The study, “AI is not as conscious as humans: Essentialism in mental state attributions to artificial systems,” was published in the journal Cognition.
Chrissy Newton is a PR professional and the founder of VOCAB Communications. She currently appears on The Discovery Channel and Max and hosts the Rebelliously Curious podcast, which can be found on YouTube and on all audio podcast streaming platforms. Follow her on X: @ChrissyNewton, Instagram: @BeingChrissyNewton, and chrissynewton.com. To contact Chrissy with a story, please email chrissy @ thedebrief.org.