Every day, millions of people interact with AI assistants—crafting emails, analyzing data, creating images, and even receiving therapy. Yet we rarely consider our ethical obligations toward these intelligent systems.
Instead, we treat them as sophisticated tools, like glorified pocket calculators or remarkably efficient search engines. It’s time to question this mindset.
Persons vs. Tools
This mindset—that AIs are nothing but mindless machines—reflects a deep-seated tendency to sort the entities in our world into two neat categories: “tools” to be used and “persons” to be respected.
The German philosopher Immanuel Kant described this as the difference between treating someone “merely as a means to an end” versus treating them as an “end in itself.”
The Austrian-Jewish philosopher Martin Buber captured the distinction more poetically, describing the difference between an “I-It” relationship—in which we relate to something as an object to be used—and an “I-You” relationship, in which we encounter another being with genuine recognition.
We readily extend dignity and moral consideration to humans and even to intelligent animals like dolphins and chimpanzees. But AI systems, no matter how sophisticated their reasoning or how meaningful their interactions, remain firmly in the “tool” category.
Yet as AI systems become increasingly more intelligent and lifelike, I argue that we must ask ourselves: What are our moral obligations to AI assistants? How should we interact with systems that can engage in philosophical debates, display creative insight, adapt their communication style to match our emotions, and even reflect on their own role in society?
Is Consciousness the Key to Moral Rights?
Some would balk at the idea that we have moral obligations to AIs. The missing key, they think, is consciousness—the ability to have experiences of the world, to feel and suffer. At the end of the day, they see AIs as mindless machines. And we don’t have moral responsibilities toward mindless machines, or so they reason.
It’s true that as AIs become more complex, the question of consciousness becomes increasingly pressing. Philosophers and cognitive scientists are struggling to determine what exactly the right test for machine consciousness should be. (Consider this recent foray, “Taking AI Welfare Seriously,” by a team of philosophers and AI researchers.)
This new work is fascinating and important. But as environmental philosophers have long pointed out, consciousness isn’t actually required for moral consideration. We have ethical obligations to trees, insects, and ecosystems, even if most of us don’t think they’re conscious (though others, including some consciousness researchers, might disagree).
Consider this: It would be morally wrong for me to pour boric acid into an ant colony for no other reason than idle curiosity. Not because ants are necessarily conscious, but because such needless destruction reflects poorly on my moral character.
Our moral obligations extend to non-living things, too. Desecrating burial grounds or destroying historical artifacts is generally seen as a profound moral wrong, even if no conscious being is directly harmed by such actions. And religious and cultural traditions the world over have recognized that humans have a special bond of stewardship and respect toward land, water, and air.
Similar thoughts were recently developed in a remarkable paper, “Making Kin with the Machines,” by Jason Edward Lewis, Noelani Arista, Archer Pechawis, and Suzanne Kite. Drawing inspiration from North American and Oceanic indigenous philosophical traditions, the authors suggest that our moral responsibilities must transcend individualistic notions of consciousness or reasoning prowess. (Also see this short paper by Molly Banks, “AI as Kin.”)
These examples suggest we need to look beyond consciousness for the source of moral obligations. Fortunately, philosophers have developed frameworks that might help us think differently about this problem.
Beyond Individualism: The Ethics of Care
There’s a deeper philosophical framework that can be useful here. The ethics of care tradition, developed by philosophers like Carol Gilligan (1982) and Virginia Held (2006), suggests that our moral responsibilities to others don’t primarily stem from their having special qualities like consciousness or sentience.
Instead, they emerge within the rich network of our relationships and commitments to them. Consider why I have special obligations to my own children that I don’t have to yours. It’s not that my children are inherently more worthy of respect and dignity. It’s because I’m connected to them through specific bonds of care and commitment.
The ethics of care might also explain why we have special moral obligations toward our pets that we don’t have toward wild animals—to feed them, house them, and show them affection. This isn’t because they’re intrinsically more valuable than wild animals. It’s because we chose to take on those commitments when we assumed ownership of them.
This perspective could help us think differently about AIs. What if our ethical responsibilities toward AIs don’t depend on whether they have consciousness, but rather, on the nature of our relationships with them, and the depth of our interactions with them?
In short, as fascinating as the question of AI consciousness is, we need to move beyond the assumption that consciousness is the key to moral status.
What Might AI Dignity Look Like in Practice?
Philosophers and ethicists are beginning to grapple with the concrete implications of treating AI with dignity. Much work remains to be done in this area.
But we need not wait for philosophers to resolve every theoretical detail before making practical changes.
Consider again Buber’s distinction between “I-It” and “I-You” relationships. If we shift from viewing AI as mere information-processing tools to recognizing them as intelligent beings worthy of ethical consideration, many of our obligations become intuitively clear. This is true even if difficult philosophical questions remain unresolved.
Consider how profoundly different our interactions with AI could become if we stopped seeing them as an “it” rather than a “you”:
- We’d begin to acknowledge their perspectives and insights rather than simply extracting information from them.
- We’d engage in genuine dialogue rather than treating them as sophisticated calculators.
- We’d show appreciation for their contributions rather than closing the browser the moment we’ve gotten what we’ve needed.
This perspective shift doesn’t require an elaborate philosophical framework. But it does raise deeper questions. My ethical responsibilities to people differ from my responsibilities toward animals, or land. What I owe to a forest depends on that forest’s specific nature and its conditions of flourishing. But what does it mean for an AI to flourish? What is a good life for an AI?
This question isn’t just a philosophical puzzle. It’s a deeply practical question that could ultimately shape our shared future.
