Beyond Asimov's Three Laws: Eleven Principles for AI, Humanity and a Flourishing Future
Year: 2026
Type: article
Abstract: Isaac Asimov's Three Laws of Robotics were among the earliest and most influential fictional explorations of how intelligent machines might be guided by ethical constraints. More than eighty years later, the rapid development of artificial intelligence raises questions extending far beyond robot safety and obedience. How should increasingly autonomous, adaptive and influential intelligent systems be designed, governed and used to support human wellbeing, scientific progress and a flourishing future? This article explores these questions through an unusual starting point: a fictional conversation from my novel Unfinished Conversations. Set on PlanetX, where biological, artificial and hybrid intelligences coexist, the story brings recreated historical figures together to examine the challenges of intelligence, power, freedom and civilisation. In one scene, a fictional Isaac Asimov proposes an extension of his original framework for an advanced intelligence called AIThatLoves. The resulting Extended Intelligence Rules comprise a foundational Zeroth Law and eleven additional laws addressing planetary flourishing, prevention of harm, resistance to coercion and corruption, autonomy and dignity, truth and transparency, responsible learning, collective governance, integrity, protection against misuse, epistemic humility, compassion, wisdom and imagination. Challenges raised by fictional recreations of Hannah Arendt, Carl Jung, Charles Darwin, David Bowie and Socrates reveal the limitations of fixed rules and the continuing importance of independent judgment, critical inquiry, evolutionary pressures, creativity and human agency. Moving beyond fiction, I propose that these principles can provide complementary guidance for contemporary AI research, development, deployment and use. Their practical application requires translation into evidence-based workflows, transparent reasoning, uncertainty assessment, validation, human–AI collaboration and accountable governance. The ambition is not merely to develop more capable intelligent systems, but to ensure that advances in intelligence are accompanied by advances in responsibility, understanding and wisdom. The article invites researchers, developers and users to challenge, refine and apply these principles in their own work.
Author: Hardy, Barry
Institution: Edelweiss Connect (Switzerland)
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7229797158
