Consciousness Commitments in Research, Policy, and Design: the LUCID Framework for Locating Underlying Commitments and Implicit Definitions

Abstract

Perceptions of machine consciousness are evolving. Conversational and autonomous agents routinely express preferences and pursue goals in manners often perceived as indicating consciousness. Developers report models with complex goal formation and apparent volitional behaviours. Whether this reflects genuine consciousness, convincing mimicry, or engineered artefacts is contested. This uncertainty is compounded without a consensus definition of consciousness. Neuroscience literature catalogues more than twenty theories, and philosophical positions expand the landscape further. HCI research routinely engages questions of machine consciousness without stating theoretical commitments, even as consequential decisions in regulation and design follow. Without explicit disclosure, theories of consciousness are invisibly and inconsistently embedded in products, laws, and public discourse. We address this gap through a dimensional framework characterising families of consciousness theory by core commitments and disagreements. Our framework diagnoses implicit positions across research, regulation, and design, revealing contradictions that produce incoherent policy and systems that simultaneously evoke and deny machine consciousness.