AI Consciousness and the Law: Towards an AI Consciousness Legal Protocol

Abstract

Artificial intelligence has not been shown to be conscious, yet the question has moved from philosophy into legislation and corporate research. Some legislatures have begun to answer it by declaration, while developers now study model welfare and the internal states of their systems. This paper argues that law can neither wait for science to settle the question nor settle it by decree. It proposes an AI Consciousness Legal Protocol: a ten-article instrument that sets the technical and legal parameters to be met before law recognises the consciousness of an AI system, entrusts that finding to an independent recognition body, and fixes the rules that govern AI in the meantime. The paper distinguishes the dictionary, scientific and spiritual meanings of consciousness, and separates the contextual awareness that AI systems already display from the valenced experience and self-directed decision that would alter legal relations. Drawing on the indicator-property method of Butlin, Long and co-authors (2023), it proposes seven legal parameters (independent evidence, valenced experience, persistent self, volition, understanding of consequences, absence of engineered mimicry, and reproducibility subject to review) and a graded four-category scale of recognition. A comparative survey of India, the United Kingdom, the European Union and the United States shows that no legal system defines consciousness as such, and that Indian law already separates personhood from consciousness. The central position is that recognition of consciousness, criminal capacity and legal personhood are three separate questions, and that human accountability for AI survives all three. The AI Consciousness Legal Protocol is a scholarly proposal offered for the consideration of lawmakers and jurists. It is not binding law and has not been adopted by any legislature or institution. Contents: 1. AI consciousness and the challenge before AI law; 2. What is consciousness?; 3. Awareness is not consciousness; 4. The technical parameters of consciousness; 5. The legal parameters of consciousness; 6. What existing law already says about consciousness; 7. The Pandora's box: mens rea, liability and personhood; 8. The AI Consciousness Legal Protocol (Articles 1 to 10); 9. Conclusion.