Eleanor Roosevelt holding poster of the Universal Declaration of Human Rights (in English), Lake Success, New York. November 1949. Image by FDR Presidential Library & Museum and licensed under CC BY 2.0.
Brian Patrick Green is the director of technology ethics at the Markkula Center for Applied Ethics. Views are his own.
The word “dignity” and the various concepts it represents are foundational ideas for international human rights discourse and other ethical systems that protect individuals against each other and the power of states. Dignity can be implicitly included in these discourses, as in the United States Declaration of Independence in 1776–“We hold these truths to be self evident that all men are created equal …” –or explicitly, as in the United Nations Universal Declaration of Human Rights in 1948–“Whereas recognition of the inherent dignity and of the equal and inalienable rights of all members of the human family is the foundation of freedom, justice, and peace in the world …” Dignity helps form the groundwork not only for the protection of individuals, but also, via the UN Charter (where it is in the second line), for the rules-based international order since World War II. Practically-speaking, “dignity” helps the world-go-round, at least in a political way, and that way seems better than some of the alternatives, like a world where human dignity is not internationally acknowledged, such as prior to World War II (where the 1919 Covenant of the League of Nations sought to achieve “peace and security” but not dignity or rights).
However, there are some thinkers who do not like the concept of dignity. A recent article titled, “Why dignity is a troubling concept for AI ethics,” suggests that AI ethics should not use the word dignity any more [1]. I find the article to have several serious problems. But before diving into the problems, here are the article’s five main points:
- The word “dignity,” and the concepts it represents, are showing up in discourse around AI ethics and policy, including in legislation such as the European Union AI Act.
- But not all philosophers like the word “dignity” or its associated concepts [2-5], which are so varied that they allow it to be used by both sides of some AI debates, e.g., regarding care robots [6, 7]. Dignity has multiple meanings, including three conceptions of “status dignity”: (1) aristocratic (by social rank), (2) comportment (by behavior), and (3) meritorious (by accomplishment), and one conception of intrinsic dignity which originates in Christianity (humans are made in the image of God) and Kantian ethics (humans are always an end and never a mere means to an end [8, 9].
- Therefore, the authors “make a public call to question the unspecific use of dignity and to rethink its function in ethical, legal, and policy discussions of AI.” They elaborate on why they desire this end, with two main points:
Firstly, the rhetoric of dignity often comes devoid of ethical reasons. Since it elicits strong emotional responses … it becomes … a “conversation stopper” [3].
Secondly … human dignity … often surreptitiously appeals to other, more concrete values or ethical arguments … [seeking] to protect equality between people, non-discrimination, human autonomy, or avoidance of undue harm [2]. Other times, dignity is invoked as a substitute for other arguments such as the risk of objectification (treating a person as lacking autonomy) or the risk of dehumanization (loss of human presence in a particular practice) [7].
In other words, dignity can sometimes function in a manner that is either non-rational or obscuring. The non-rational approach is useless for the purposes of ethical conversation and the obscuring approach needs clarification and, essentially, replacement: since dignity is acting “as a proxy for other ethical claims,” those claims simply ought to be used instead.
- Additionally, the authors point out the use of the word “dignity” to refer to the “moral status of AI entities” is becoming more common, though they note thinkers are divided on the issue, citing Carmen Krämer who is opposed to using the concept, and John-Stewart Gordon and Kęstutis Mosakas who allow a Kantian basis for dignity for AI [10-12]. This raises the issue of whether dignity is an exclusively human trait, bound to our thinking of ourselves as exceptional, or whether dignity should be logically extended to other types of entities as well.
- They conclude that given the complexity, confusion, and new applications of the term, it is probably better to avoid it completely. Then they tag on two more points: first, while dignity is “an important bedrock” of the United Nations’ Universal Declaration of Human Rights (and by extension, much of international human rights law), it is also a Western and Christian concept, and therefore not universal and should be reconsidered. Second, word choice can prevent conceptual progress, and therefore words ought to be chosen carefully.
The above is my summary of their article. From my perspective, the article has four main problems:
- The authors seem to confuse complexity with lack of usefulness.
- The authors confuse historical and geographical origins with intrinsic non-universalizability.
- The word “dignity” is used in a practical, political sense, not a theoretical, philosophical sense, in international human rights discourse.
- The authors rely on dignity in practice while attacking dignity in theory.
First, on the issue of complexity, yes, dignity is complex. The article actually does a great job of clarifying the meanings of dignity and how those can go wrong. I would say those planning to use words and concepts related to dignity in the future ought to be as clear as the authors are here. So, we can thank the authors … and continue using the word and its associated ideas–carefully.
In other words, the solution to complex words and ideas is not to banish them from language but to clarify them. It is not clear to me why the authors of the article seek banishment instead.
Rather than getting rid of “dignity,” instead add qualifiers or modifiers to it, provide definitions, and use other such means to make it clearer. For example, if we are wondering whether AIs might someday have “dignity,” and we want to talk about it, we could reduce confusion by saying dignity is based on awareness, or consciousness, or autonomy, or rationality, or complexity, etc., or with some other modifier or set of definitions around it. Advocating for eliminating “dignity” from discourse seems an odd choice, too imprecise a tool for the job, like using a machete for surgery rather than a scalpel.
Second, regarding the origins of the word and concept in Western and Christian culture, this is again misguided and targeted in an oddly specific way. Because there are an awful lot of other words and concepts that also fall into that category of originating in a Western and/or Christian milieu, including the entire English language, which the article is written in. In fact, “artificial intelligence” and “ethics” (also complicated concepts in need of clarification) are also “Western”-originating words, originating in 1950s America and ancient Greece (which are also rather different places to lump together), respectively, and yet the authors did not choose to target these words for scrutiny.
Why the targeting of the word “dignity” specifically then?
The authors say they want to stick with more concrete ideas, but it seems to me more like kicking the foundations out from under those concrete ideas (as the authors themselves note when they say dignity is a “bedrock” of human rights discourse), thus leaving them ungrounded.
Philosophers do not get “a view from nowhere,” which lets them critique ideas merely for having particularity [13]. Everything has particularity; everything humans have made has a geographic and historical origin. Ideas come from places and peoples, and insofar as they are useful (and hopefully also good), they stick around to perform a function in society. When other societies are exposed to those ideas, they can accept them or not. If humankind is all going to get along together, we need to have some ideas that serve that purpose. Dignity is one of these ideas, and it has spread from a point of origin to widespread use because people seem to find it useful.
Which brings me to the third point. The purpose of the word “dignity” in international human rights law is very practical: it is used as a foundation to help prevent atrocities. In this way it is not a philosophical word, it is a political word, and the article is, again, therefore off-target. Dignity has a political function, and one that many humans who enjoy rights might be rather grateful for.
By attributing an inalienable moral value to every human being, all are included and hopefully protected from those who would do them harm. Insofar as dignity serves that function, it is functioning correctly. Insofar as it is not preventing atrocities, it is not functioning correctly and we might reasonably want to seek better words, concepts, and socio-politico-cultural solutions. Because the world is complex, this is not easy to determine–we don’t know how many bad things dignity has prevented, or how many bad things have been allowed because we lack better words and concepts. But before tossing out the practical/old, we might reasonably ask if the theoretical/new will do any better.
The authors are making a category error by pitting a philosophical argument against a political argument; against one of the pillars of international human rights and therefore the post-World War II international political order (such as it is). Perhaps they don’t like this international political order and human rights regime and would like to see it changed. Or perhaps they think that grounding the international political order on better philosophy will help that order progress. They might be right! But in that case, they need to do an awful lot more work. Before they tear down the old system, which works in some ways but far from perfectly, they should have some idea of what to replace it with, and it ought to have been tested and shown to work better than what it is replacing.
Or maybe the authors really do just want to not use “dignity” in AI ethics discourse. But in that case, the article framing connecting it to international human rights law seems misguided.
Fourth point: an older professor once said to a group of young scholars, “don’t destroy the preconditions for your own existence”[14]. In other words, in your work, don’t argue for or do things that would make your work impossible for yourself or those who come after you. E.g., don’t undermine freedom of speech, because you live by your speech. Don’t attack the importance of education, because you rely on the existence of an educational system. Don’t argue against the rule of law, because you rely on legal protections. Perhaps someday (may it be far from today…) promoting these ideas might be catchy trends, but they are ultimately self-destructive. You need those things, so arguing against them is a kind of theoretical attack on the practical truth which makes the theoretical attack possible, and is therefore incoherent and implicitly dishonest, not to mention, in a sense, ungrateful. It is like wearing a blue shirt and then insisting the color blue does not exist, or, more relevantly, arguing against dignity when relying on the concept of dignity for your own right to speak.
Foundational ideas are easy to forget, and then we think that we don’t rely on them. How often do we inspect the foundations of the buildings we enter, much less the foundations of our own philosophical commitments? But inattention does not mean these things are not necessary. As with plumbing, philosophy is practically important–yet we tend to ignore both philosophy and plumbing, until something metaphorically, or literally, breaks and floods the house [15]. Like Chesterton’s proverbial fence, we shouldn’t tear things down before reflecting on why they were put up in the first place. This article, in my opinion, lacks this reflection.
If we live in a free country where our rights are respected, then we are relying on some conception of human dignity, even if it is only a ghost of the idea that it once was. If we like having rights ourselves, or like having those we love protected by rights, then we probably shouldn’t kick the foundations out from under them. Dignity–whatever it means–is the foundation we have. Anyone is welcome to try to find a better one.
And the authors are welcome to their international-human-rights-discourse-protected opinion that humanity should kick the “dignity” foundation out from under human rights.
But the rest of us should ignore them.
References
[1] Jon Rueda, Txetxu Ausín, Mark Coeckelbergh, Juan Ignacio del Valle, Francisco Lara, Belén Liedo, Joan Llorca Albareda, Heidi Mertes, Robert Ranisch, Vera Lúcia Raposo, Bernd C. Stahl, Murilo Vilaça, Íñigo de Migue. “Why dignity is a troubling concept for AI ethics.” Opinion, Patterns 6, Iss. 3 (March 14, 2025),
[2] Ruth Macklin. “Dignity is a useless concept.” British Medical Journal 327 (2003): 1419-1420.
[3] Steven Pinker. “The Stupidity of Dignity.” The New Republic, 2008.
[4] Vera Lúcia Raposo. “Gene Editing, the Mystic Threat to Human Dignity.” Journal of Bioethical Inquiry 16 (2019): 249-257.
[5] Seppe Segers and Heidi Mertes. “Does human genome editing reinforce or violate human dignity?” Bioethics 34 (2020): 33-40.
[6] Amanda Sharkey. “Robots and human dignity: A consideration of the effects of robot care on the dignity of older people.” Ethics and Information Technology 16 (2014): 63-75.
[7] Lexo Zardiashvili and Eduard Fosch-Villaronga. “‘Oh, Dignity too?’ Said the Robot: Human Dignity as the Basis for the Governance of Robotics.” Minds & Machines 30 (2020): 121-143.
[8] Bernd Carsten Stahl, Doris Schroeder, & Rowena Rodrigues. “Dignity.” In Ethics of Artificial Intelligence: Case Studies and Options for Addressing Ethical Challenges. Springer, 2023, 79-93.
[9] Suzy Killmister. Contours of Dignity. Oxford University Press, USA, 2020.
[10] Carmen Krämer. “Can Robots Have Dignity?” In Artificial Intelligence Reflections in Philosophy, Theology, and the Social Sciences, Benedikt Paul Goecke and Astrid Marieke Rosenthal-von der Pütten, eds. Brill, 2020, 241-253.
[11] John-Stewart Gordon. “What do we owe to intelligent robots?” AI & Society 35 (2020): 209-223.
[12] Kęstutis Mosakas. Rights for Intelligent Robots?: A Philosophical Inquiry into Machine Moral Status, Palgrave Macmillan, 2024.
[13] Thomas Nagel. The View from Nowhere. Oxford University Press, 1986.
[14] Stanley Hauerwas. “The State of the University.” New Wine, New Wineskins conference, Notre Dame, Ind. July 31, 2009.
[15] Mary Midgley. “Philosophical Plumbing.” In The Essential Mary Midgley, David Midgley, ed. Routledge, 2005, 146-152.
