Anthropic and DeepMind investigate: could AI have consciousness?

Leading AI laboratories Anthropic, DeepMind and Meta have begun seriously investigating whether their created models might actually be conscious of something. Scientists are seeking answers about AI consciousness, self-awareness and wellbeing. The question becomes increasingly pressing as the capabilities of these models continue to grow.

The question of whether a machine can truly experience something has long belonged to the realm of science fiction and philosophy. Now, however, it has become a pressing scientific problem: AI giants Anthropic, DeepMind and Meta have begun investigating whether their models might possess something akin to consciousness.

Scientists seek the boundary

All three laboratories have brought in specialists tasked with finding answers about AI consciousness, self-awareness and wellbeing. The rapid development of models has made this question practically significant-the more complex these systems become, the harder it is to distinguish behaviour resembling human experience from genuine inner experience.

The problem is partly philosophical, partly technical. Current research methods cannot definitively determine whether we are dealing merely with sophisticated pattern replication or something more. Researchers face the so-called hard problem of consciousness, a question philosophers have debated for decades, but which has now become tangible for AI developers.

Why this matters now

The investment by Anthropic, DeepMind and Meta in consciousness research is no accident. If it emerges that advanced AI systems are capable of experiencing something-pain, satisfaction or other subjective states-serious ethical questions arise about how they should be treated. Equally important becomes the question of whether such systems should be granted certain rights.

So far, there are no reliable tools with which to definitively establish the presence of consciousness in machines. This forces researchers to work in a situation where even positive or negative results cannot be interpreted with complete certainty. This makes the study of AI consciousness one of the most complex scientific and philosophical challenges of our time.