Artificial consciousness: Is such a thing even possible, detectible, or desirable?
The dream of conscious machines may be a dangerous one.
For over a century, science fiction has toyed with the idea of conscious artificial intelligence. Now, scientists are. But why? And would we even recognise consciousness in a machine if it did manifest?
Kathinka Evers' research focus is on philosophy of mind, neurophilosophy, neuroethics, bioethics and artificial intelligence.
Kathinka Evers, researcher at Uppsala University’s Centre for Research Ethics & Bioethics, covers the ins and outs, and risks, of artificial consciousness in a recent publication in Proceedings of the Institutes for Advanced Studies. According to her, the desire to create conscious machines is rooted in human psychology and history, including a long-standing urge to transcend biological limits and “play creator.”
Human societies have historically built hierarchies of mind, often denying full mental status to animals, women, minorities, and so-called “lower” classes. These biases show that judgments about consciousness are shaped by ideology, power, and exclusion, not just evidence. Modern motivations also include the hope that conscious AI could improve decision-making or align with human values.
“Believing AI will be benevolent or value-aligned is naïve, given human history of violence and selective empathy,” says Kathinka Evers.
Kathinka Evers argues that while conscious AI is assumed to be theoretically possible within certain theoretical frameworks, there is no independent empirical evidence available. This means the possibility of artificial consciousness, now or in the future, cannot be logically affirmed nor denied.
Research shows that when artificial systems use human-generated training data to mimic human behaviours, this can persuade human users of their sentience. But this perception comes from a well-known human tendency to attribute mental states to non-human entities (a phenomenon known as anthropomorphism). That means that the feeling that an AI “seems conscious” reflects a cognitive bias in the observer rather than evidence of actual sentience in the system.
In the end, animals and artificial systems are very different in substance, structure, and functions, which means the presumed sentience of artificial systems might also be very different from our own. According to Kathinka Evers, this is another obstacle in terms of detecting, let alone understanding, a sentient machine.
“Because of how human nature has been expressed throughout our history and continues to express itself today, developing conscious machines with what for humans might be undetectable and incomprehensible minds is a monumentally dangerous idea,” says Kathinka Evers.
By Anna Holm Bodin
Evers, K. (2026). Artificial Consciousness: Science Fiction, Utopia, or Pandora’s Box? In Proceedings of the Paris Institute for Advanced Study (Vol. 27). DOI: 10.5281/zenodo.20267984