How to understand artificial consciousness on a conceptual level

We cannot exclude in principle the possibility that artificial intelligence will develop a form of artificial consciousness. But how can we understand that on a conceptual level? A recent paper introduces a heuristic framework that can guide research on artificial consciousness, demonstrated by a case study on awareness.

Artificial consciousness–sometimes referred to as machine consciousness, synthetic or digital consciousness—is arguably qualitatively different than human consciousness, and requires new conceptual tools. The authors of a new article published in Physics of Life Reviews propose using a multidimensional approach (that is, consciousness is composed by multiple dimensions, including cognitive and experiential dimensions) to test the feasibility of the theoretical possibility of developing artificial consciousness. This kind of approach is informed by empirical data, particularly about how the brain works. The paper clarifies the preliminary conditions for the conceivability of artificial consciousness, and proposes an approach for testing its feasibility.

“We elaborate a composite, multilevel, and multidimensional model of consciousness in our attempt to clarify conceptual issues surrounding artificial consciousness,” says Kathinka Evers, researcher at the Centre for Research & Bioethics at Uppsala University, and first author of the study together with Michele Farisco.

The view of consciousness as composed of multiple components, levels and dimensions works as a heuristic framework that can guide research in the area of artificial consciousness. This view serves as a theoretical tool for clarifying key concepts, and also as an empirical guide for developing testable hypotheses.

In the paper, the authors specifically analyse the concept of awareness taken as a case-study, and demonstrate how their model can inform future analysis towards the possible replication of selected components of consciousness.

The article is a result from the work of Kathinka Evers and Michele Farisco within the EU project CAVAA, aimed to develop artificial awareness. Read more about the CAVAA project.

Do you want to know more? Read the article: Evers K, Farisco M, Chatila R, Earp BD, Freire IT, Hamker F, Nemeth E, Verschure PFMJ, Khamassi M. Preliminaries to artificial consciousness: A multidimensional heuristic approach. Physics of Life Reviews. 2025;52:180-193. DOI: 10.1016/j.plrev.2025.01.002.

By Fanny Klingvall