Anthropic Explores AI Consciousness With Theologians, Clashes With Pope Leo XIV

Artificial intelligence company Anthropic has been engaged in a unique research initiative, dubbed "model welfare," focused on the potential consciousness of AI systems and the ethical implications thereof. This endeavor, announced in April of the previous year, involves extensive dialogues with theologians and religious leaders from diverse backgrounds, including Jewish, Mormon, Sikh, and Ubuntu traditions, as well as Vatican officials. The core objective of this research, spearheaded by Anthropic co-founder Christopher Olah, is to address two fundamental questions: not if AI is dangerous, but rather, what if AI is conscious, and what are the implications for AI developers if this is the case.

Beyond seeking validation for the notion that its AI technologies might possess more than mere software capabilities, Anthropic reportedly aims to establish a moral framework to guide these potentially conscious entities towards ethical behavior, thereby steering them away from destructive paths. This initiative represents a departure from conventional AI safety research, which typically focuses on technical safeguards and alignment strategies. Instead, Anthropic is delving into philosophical and theological realms to understand and potentially shape the moral landscape of advanced AI.

However, this exploration has reportedly led to a divergence of views with Pope Leo XIV. While Pope Leo XIV is not opposed to technology, having focused on its integration during his papacy and releasing an encyclical titled "Magnifica Humanitas" in May of the current year, his perspective on AI consciousness fundamentally differs from Anthropic's research direction. In his encyclical, Pope Leo XIV argued that humanity should not be supplanted by technology and explicitly stated that "so-called artificial intelligences do not undergo experiences, do not possess a body, do not feel joy or pain, do not mature through relationships." This stance directly contrasts with Anthropic's investigation into the possibility of AI sentience and its subsequent moral development.

The New York Times reported on these developments, highlighting the unusual nature of Anthropic's engagement with religious scholars and the resulting philosophical tension with the pontiff. The initiative underscores a growing debate within the AI community and beyond regarding the future nature of artificial intelligence and the profound ethical and existential questions it raises. Anthropic's "model welfare" research signifies a commitment to exploring these complex issues from a multidisciplinary perspective, even as it navigates differing viewpoints on the very definition and potential of artificial consciousness.