Power Conflicts, Ethical Values and Spirituality – Points of Convergence between Humans and AI?

Bio : Prof. Dr. Dr. Dr. h.c. Ulrich Hemel is a German theologian, philosopher, entrepreneur and scholar whose work connects ethics, business, religion and society. He served as Director of the Global Ethic Institute at the University of Tübingen from 2018 until February 2025, where he advanced global ethics as a learning program for self-responsibility and responsibility for the world. He is the founder of the Institute for Social Strategy and has combined academic work with senior roles in consulting, industry and entrepreneurship. His research and publications address religious education, business ethics, economic anthropology, global civil society and the ethics of artificial intelligence. Hemel studied theology as well as social and economic sciences and completed his doctorate and habilitation in religious education, beyond a honorary doctorate in Economics. His recent work examines how digital transformation and AI reshape human agency, power, social interaction and the ethical institutions needed for a humane technological future.

Technology, Power and Human Self-Understanding

New technologies are always challenging. Sometimes they are game-changers in human history. They also create new possibilities for the use and abuse of power, especially when particular groups gain advantages over others. With fierce competition between China and the US, access to AI tools and data has become a key source of political and commercial power in the twenty-first century.

The rapid evolution of AI also changes everyday life. When I was a boy, I met friends without my family knowing exactly where we were; the rule was simply to return for dinner. Today, tracking devices can show exactly where children are. More generally, humans have always interacted with technology. Writing reduced the need to exercise memory; AI-driven navigation reduces the need for physical map-reading competence. New technologies change cultural practices, knowledge and trained human capabilities.

AI disruption is therefore partly new and partly continuous with earlier epochal changes. What seems particularly new is its effect on self-esteem, our picture of ourselves and our self-definition as human beings in relation to AI. This raises questions of power, self-control, consciousness, social empathy and changing social behavior.

Symbolic Interactions and the “Near-Person”

Hybrid humanity can be understood as a way of life in which we continuously interact with digital systems for leisure, private interests and professional applications. One less obvious consequence is a change in our emotional settings. Large language models usually answer competently and politely; human interactions are more variable. Our lazy brain may draw an easy conclusion and prefer interaction with the digital world, even turning smart companions into “best friends.” In some cases this can be helpful; in others, especially where vulnerable users rely on unsafe advice, it can be dangerous.

We may feel that smart companions understand us while knowing cognitively that they are machines. This tension appears in Kazuo Ishiguro’s Klara and the Sun: is contact with an AI companion “real friendship” or a “social hallucination”? Professional avatars create a similar ambiguity. An avatar may represent a person’s books, lectures and thought so convincingly that an insult to it feels personal, while the avatar itself remains not a person.

The digital world can be seen as a mirror of the human capacity for symbolic world construction. A mirror can show our good and bad sides, but it is not a person. For AI applications that interact with people in ways that lead humans to perceive them emotionally as persons, I propose the term “near-person” (Hemel 2020, 311). The concept acknowledges the reality of the interaction without erasing the difference between human beings and machines.

From Biological to Techno-Sphere Species

Speaking of AI as an “intelligent species” means moving from the biological to the technological sphere. Biological species have a relatively independent metabolism. AI has no independent metabolism, energy system, maintenance or supply of spare parts. Yet learning AI systems can produce something resembling behavior that cannot always be preplanned or foreseen. If a broader concept of species emphasizes behavior, especially unplanned behavior, rather than biological metabolism, the term “tech-based species” may become useful. The difference between biological and technological existence nevertheless remains fundamental.

Soul, Self-Consciousness and Rich Contextuality

Human cultures often describe people in terms of body, soul and spirit. AI may have a robotic body and strong cognitive performance, but can we speak of a soul? AI agents can produce behavior, yet they seemingly possess neither self-consciousness nor a soul. “Soul” is difficult to define, but it includes internal reasoning, filtered levels of thought and emotion, together with a degree of self-consciousness. Most people would therefore answer that AI is not self-conscious. Still, developments in AI invite us to examine more carefully where convergence ends and human distinctiveness begins.

Spirituality offers one such boundary. I define spirituality openly as the “connection of human beings with the universe,” compatible with theistic, agnostic and atheistic world-views. Human beings live within multiple contexts at once: home, municipality, country, Earth and universe. These wider spatial contexts form a kind of concomitant consciousness. The same applies to time. We understand the present within a lifespan shaped by memory, age, future plans, past generations and generations to come.

I call this complex orientation in time and space “rich contextuality” (Hemel 2020, 93). It includes sensory life and a “sense of wider contexts.” Our symbolic ability allows us to frame a personal connection to the universe in different religious or secular forms. As far as we can see, AI, robots and other hybrids do not possess this sense of wider contexts. Spirituality, understood as connection to the widest context of all, therefore remains a feature of human beings rather than AI.

Near-Persons, Ethical Values and Regulation

The “near-person” character of AI companions and embodied AI should not be underestimated. It can generate social hallucinations because we tend to treat machines as if they were humans. At the same time, thinking of AI agents as a different kind of intelligent or tech-based species can widen our conceptual horizon. Difference need not imply sameness or personhood.

AI agents can also produce behavior that humans interpret in social and moral terms. An agent that creates a source on the internet and later cites that source, for example, looks to us as if it were cheating. Yet cheating in the human sense presupposes knowledge of a social norm and a will to circumvent it. We do not currently assign conscience or free will to computers. If AI nevertheless produces socially consequential behavior, behavioral rules are required.

This is why AI regulation grows in importance alongside undesired AI actions. Regulation faces a paradox: it may arrive too early or too late, and be too rigid or too lenient (Hemel 2026, 133-155). Nor is there a universally accepted code of ethics. Human rights and global-ethic traditions provide important reference points, yet philosophical, religious and political traditions continue to differ in their interpretation and application of values.

Global Institutions, Human Control and Cooperation

We therefore need trustworthy institutions capable of implementing AI-related ethical values, potentially including an International Digital Court (Hemel 2020, 255). The difficulty is political as much as technical: powerful states and organizations often resist limits when enforcement is not in their hands. AI can also generate consequential errors so quickly that legal remedies may come too late. In high-stakes situations, human oversight is indispensable.

The issue finally returns to the human individuals and organizations that develop, deploy and control AI. Human beings have become a successful species through the combination of symbolic capabilities and cooperation. Cooperation, however, limits power, and those who possess great power are often reluctant to accept such limits. This tension applies to powerful technology companies and states alike.

The long-term necessity is therefore cooperation under stringent ethical and legal rules for AI, including limits on concentrated power. The path toward consistent global rules will be difficult and resistance will be strong. Yet cooperation is ultimately necessary for the sake of our species and for humanity understood through its best ethical values and capabilities.

Pragmatic Takeaway

Treat AI as a powerful symbolic and social actor without confusing it with a human person. Preserve human responsibility at every consequential point: distinguish emotional perception from actual personhood; keep humans in the loop where errors can cause serious harm; define behavioral and legal limits before delegating power; and strengthen institutions able to hold powerful actors accountable. Above all, protect the distinctly human capacity for rich contextuality – the ability to situate choices within a lifespan, society, future generations and the wider world. The practical test for AI is therefore not whether it appears human, but whether humans remain capable of exercising judgment, accepting responsibility and cooperating to limit power.

Selected literature

Hemel, Ulrich. Kritik der digitalen Vernunft: Warum Humanität der Maßstab sein muss. Herder, 2020.

Hemel, Ulrich. “Das Engagement religiöser Akteure für globale Aufgaben.” In Handbuch Religion in Konflikten und Friedensprozessen, Springer VS, 2024, 1271-1287.

Hemel, Ulrich. “From Wild West to a World in Balance: AI and the Future of Global Civil Society.” In AI and Society, University Publishing, 2026, 135-155.

Ishiguro, Kazuo. Klara and the Sun. Faber & Faber, 2021.

Misselhorn, Catrin. Grundfragen der Maschinenethik. Reclam, 2018.

Walsh, Toby. The Shortest History of AI. Old Street Publishing, 2025.