Old argumentative plays on a new field? Contributions from sport ethics to the ethical analysis of artificial intelligence

Abstract In the ethics of artificial intelligence, a recurring anxiety concerns the deskilling argument: the worry that delegating distinctively human tasks to AI erodes valuable capacities and there by diminishes meaning. This article argues that philosophy of sport offers underutilized but illuminating resources for analyzing this concern. Drawing ondebates over biomedical enhancement in sport ethics, I identify four argumentative strategies—centered on the values of agency, effort, natural talent, and narrativity—that sport philosophers have developed to evaluate themoral costs of skill-displacing technology. These “old plays,” I contend, can be productively transplanted onto the “new field” of AI ethics, where analogous debates about skill erosion, meritocracy, difficulty, and human identityremain contested. Sports, understood as games of physical skill, constitute a privileged analytical setting: their emphasis on process over outcome makes the exercise of skill especially visible and prominent. By recovering insights from thinkers such as Bernard Suits, Scott Kretchmar, John Gleaves, Michael Sandel, and Alasdair MacIntyre,this article advances two aims: to vindicate the philosophical contribution of sport ethics as a discipline, and to enrich normative reflection on AI’s deskilling potential with arguments already tested in analogous debates. Similar content being viewed by others 1 Introduction In the debate surrounding the ethics of artificial intelligence (AI), ethical concerns about the impact of this technology are often examined by analogy. AI advocates claim that these new tools are analogous to earlier technological developments that, despite raising similar ethical anxieties, have ultimately been accepted by society and become central to people’s lives. Hence, this argumentative line concludes that AI should also be accepted.Footnote 1 AI critics, in contrast, use analogical thinking to identify commonalities between AI and prohibited or heavily regulated technologies, such as nuclear power (Hofman et al. 2023). In this article, I expand on this analogical approach by exploring a continuity between current technologies and AI: their deskilling effects. Drawing on Ferdman (2025), I define deskilling as the loss of valuable, distinctively human abilities as a consequence of the implementation of a new technology in society or, more specifically, within particular social practices. Hence, to analyze the continuity mentioned above, I focus on one social practice where deskilling concerns constantly arise: sport. Specifically, deskilling concerns have had a long and strong presence in the ethical debate on the biomedical enhancementFootnote 2 of athletic performance (e.g., Reid 2017). Thus, the theoretical insights in that academic discussion are fruitful to scholars in AI ethics. My central contribution involves recovering chief sport ethical intuitions (i.e., their “old argumentative plays”) to shed light on the deskilling debate in AI ethics (i.e., a “new field”). In particular, I pursue two aims: (1) to vindicate the value of reflections produced within sport ethics and (2) to employ them to contribute to thought about the ethical dimensions of AI (at least with respect to its deskilling potential). I proceed as follows: I first explore the deskilling debate in AI ethics to then draw parallels with the same debate in sport ethics, which is at the core of discussions on the moral permissibility of biomedical enhancement, that is, the use of pharmacological substances or medical procedures to boost athletic capacities. By examining the latter body of literature, I identify argumentative strategies in sport ethics that can shed light on efforts to address the implementation of AI in broader society. All these strategies share one element in common: they build on valuable human traits associated with skillful performance (i.e., agency, effort, natural talent, and narrativity) that contribute to developing meaningful sporting practices. 1.1 The deskilling debate in AI ethics In an examination of meaningfulness in an AI-driven world, “where AI systems partially or fully take over tasks that humans previously executed using their natural intelligence”, Rüther (2025, p. 1404) explicates that a common argumentative line holds that outsourcing human tasks to machines creates a gap between people’s interest in performing meaningful tasks and their actual ability to realize that interest, if not fully, at least partially. In essence, the argument has the following structure: - 1. Certain activities confer meaning; - 2. When outsourced to machines, those meaningful activities become unavailable to humans; - 3. Such a lack of availability leads to a loss of meaning. Different versions of the “meaning gap” view allude to diverse, valuable features that help explain the origins of such meaning. The deskilling approachFootnote 3 places the onus on human skills, understood as embodied abilities to achieve an end specific to a human domain with maximum certainty and minimal time and energy expenditure.Footnote 4 For example, taxi drivers rely on memorization skills to navigate a city and take customers to their destinations with minimal time and energy. A translator relies on their ability to switch between different languages to facilitate communication with maximum certainty and minimal expenditure. Outsourcing tasks that involve the exercise of these skills to AI deprives humans of the meaning such activity produces. For instance, with the invention of GPS systems, taxi drivers’ skillful navigation skills became obsolete. Similarly, translation tools such as Google Translate and DeepL have significantly reduced the need for translators’ expertise. Critics of the deskilling approach, as formulated so far, note that it lacks normative force: failing to exercise certain human skills, or even eliminating opportunities to do so, hardly constitutes a moral wrong. Many people lack the skills to grow or hunt their own food because they rely on others and technology to meet their nutritional needs. Yet, no one opposes the development of farming and food-processing technologies on deskilling grounds (Aylsworth and Castro 2024, p. 7). Moreover, societies promote skill erosion by emphasizing certain basic capacities (e.g., critical thinking) and encouraging individuals to overspecialize in their preferred tasks and skills. Likewise, if machines can perform a mindless, tedious task that used to be done by humans (e.g., controlling traffic during regular vehicle circulation), replacing low-skilled activities with mechanical means is a no-brainer. What occurs here is a process of reskilling, in which the reduced use of certain capacities is compensated for by increased use of other skills or by the functioning of the system in which individuals operate (Rafner et al. 2021). Hence, eliminating opportunities to exercise certain skills for some, or even most, people is morally innocuous. In other cases, eliminating such opportunities can be highly beneficial, as in upskilling processes, where individuals rely on technology to develop broader, higher-level skills. Consider a space engineer using complex computational software to complete calculations to understand the trajectory and interactions of a series of celestial objects. Therefore, what needs to be made more explicit is why skills, or at least certain skills, possess normative value, making their erosion or removal morally condemnable. Efforts to justify the value of skills are numerous and, typically, draw from Aristotelian and/or Kantian conceptual resources. From an Aristotelian perspective, the most relevant notion is “human flourishing”, that is, a human condition involving “the competent exercise of our human capacities to produce intrinsically valuable outcomes” (Sung and Ferdman 2025, p. 4). As value theorist Bradford (2015) explicates, to achieve the central aim of the activities in which they participate, humans rely on capacities whose competent exercise is usually inherently valuable. For instance, in team sports, athletes must exercise capacities in skilled movement, strategic thinking, and cooperation to accomplish the goal of the activity.Footnote 5 However, only the cultivation of those capacities identified as most typical of the species (e.g., rationality) leads to human flourishing. Hence, Sung and Ferdman (2025) identify three sources of “unflourishing”: (1) unflourishing that results from the competent exercise of a capacity that nevertheless produces a negative output; (2) unflourishing that results from the insufficient exercise of the capacity, thereby producing a negative output; and (3) unflourishing that results from the impoverishment of a capacity. (p. 4) Reliance on AI most closely connects to (2) and (3). Delegating human tasks to AI-powered technologies creates situations in which individuals have few opportunities to exercise their capacities, and, as a result, their reliance on these technologies diminishes their capacities. Take the GPS example mentioned above. By constantly relying on such technology to navigate, individuals have little need to use their own navigation skills, perhaps only in situations where GPS lacks information or stops working. In these cases, moreover, because of their lack of prior exercise and refinement, those capacities remain underdeveloped and weak. Aylsworth and Castro (2024) appeal to Kant’s notion of self-respect for one’s humanity, one’s capacity to self-govern. For the philosopher from Königsberg, individuals have a duty to respect humanity, which, in the case of their own humanity, involves resisting their own oppression, for instance, by cultivating their humanity. From this standpoint, deskilling is morally problematic when it affects the subset of capacities essential to people’s exercise of their autonomy. Delegating tasks that involve abilities such as weighing reasons, thinking through complex problems, critically assessing information, and pondering multiple life options is morally controversial because it would imply violating the duty to cultivate one’s humanity: There are many skills that we could outsource to a machine without compromising our autonomy. We can let them correct our spelling, find the best route to work, and remind us of our appointments. None of these abilities have immediate implications for your humanity. But imagine what it would be like to hand over your ability to critically assess your values, to forfeit the capacity to reflect on your conception of the good life and revise it. Such a forfeiture would amount to giving up authorship of your life story. (Aylsworth and Castro 2024, p. 17) In a similar vein, Krook (2025) argues that AI’s biggest threat to humanity is to produce situations where individuals no longer need to rely on decision-making, critical thinking, or creativity skills, those most typically human: “The biggest threat to humanity is therefore not that machines will become more like humans, but that humans will become more like machines, as more passive and mechanical entities” (p. 6011). Combining Aristotelian and Kantian elements, the capabilities approach also offers insights similar to those of the two traditions (Ratti and Graves 2025). A capability is a substantial freedom—that is, a range of feasible options available to a person for achieving a concrete outcome in the specific context in which they operate. For instance, individuals in Global North and Global South societies have different capabilities to be nourished or to engage in play. In Nussbaum’s (2013) formulation—the most widely accepted and discussed—the capabilities approach identifies a set of basic capabilities whose realization, at least to a certain threshold, is required for humans to be worthy of dignity. Among these capabilities are bodily health, practical reason, affiliation, play, and control over one’s environment. Given its potential to significantly affect people’s actions, contexts, and interactions, AI tools can have a profound impact on individuals’ substantial freedoms: “AI tools shape our own capabilities, and by doing so, shape our substantial freedoms, and hence they are ‘ethical’ even when they do not do anything ‘moral’ or ‘immoral’—just by operating, they profoundly shape the ethical dimension of our existence” (Ratti and Graves 2025, p. 6). For instance, large language models’ (LLMs) capacity to produce written essays for students erodes these individuals’ ability to exercise critical thinking (Lee et al. 2025). This skill is central to the practical reason capability, which involves being able to form a conception of the good and to engage in critical reflection on the extent to which one’s life realizes it. The deskilling processes resulting from AI implementation are not without precedent. Examining them through the lens of the biomedical enhancement debate in sport ethics offers a helpful, arguably simpler, social practice-centered framework for identifying precisely what makes the implementation of such a technology ethically problematic. From this perspective, technology’s negative impact on central capacities is morally wrong when it erodes both a valuable human trait and the purpose of a meaningful, cooperative human activity. 2 The deskilling debate in sport ethics Sport serves as a privileged window onto broader ethical and social questions (López Frías 2025). After all, despite having their own (relatively autonomous) configuration, sporting activities are heirs to their time, with larger ethical and social issues constantly permeating them.Footnote 6 An advantage of looking at sport for clarity on those issues is that it offers a simpler, clearer structure.Footnote 7 Also, modifying some of its elements to observe the effects of those alterations is more feasible than in broader social spheres.Footnote 8 As such, examining an ethical or social debate in sport can be extremely helpful in tackling the same or a closely related issue on a larger scale. Among these debates is the use of biomedical performance-enhancing substances or methods in sport, which sheds light on deskilling processes resulting from the implementation of AI, a (performance-enhancing) technology widely relied on to complete everyday tasks. Despite the differences between AI implementation and biomedical enhancement in sport, both phenomena involve replacing (and/or supplementing) skilled performance with technological means. Sports are games of physical skill (Meier 1988; Suits 1988). Games encourage participants to focus on the process of achieving a goal. For instance, in soccer, the goal is not simply to put a ball in a net, but rather to do so while abiding by the rules of the sport: using any part of the body other than the arms, cooperating with teammates, facing an opposing team, etc. The inextricable link between prescribed means (i.e., skills) and goals, according to Suits (2005), is a unique trait of games. Achieving the goal in a game is conditional on using the right means. Hence, means are more important than goals. What matters the most is the process by which one achieves the goal of the game (López Frías 2026). This emphasis on processes starkly contrasts with the focus on goals or results common across most other human activities, which have a more instrumental or, in Suits’ terms, technical character. In technical activities, goals take priority over the means, which are selected primarily for their efficiency. For instance, in carpentry or science, what matters is to achieve a goal: building a house or discovering the truth. Consequently, humans must use the most efficient means possible. Today’s carpenters seeking to build a house would be called fools if they used less efficient means (e.g., a hand drill) over more efficient ones (e.g., a power drill). The point of their activity is to build the house, not to do it in a heavily regulated manner. In games, this logic is inverted. If cyclists use more efficient means (e.g., a motorbike or a car) to complete a race, they cease to participate in the sport. The means are inextricably linked to the goal, enabling participants to develop and exercise very concrete skills. The centrality of skill exercise in games makes these activities a privileged setting for examining deskilling effects on two grounds: constitutive and normative. From a constitutive perspective, because games are defined by the requirement that goals be achieved through prescribed means (i.e., certain skills), they are uniquely vulnerable to a specific form of technological disruption: the bypassing of those means. When a biomedical enhancement substitutes for the skilled process that a game prescribes, it hollows out the activity from within. Moreover, from a normative standpoint, in games, one can appreciate both the instrumental and—especially—the intrinsic aspects of skills, the latter of which is typically overlooked in most other social activities in favor of their instrumental features. Games are better suited than other activities to promote autotelic engagement—that is, engagement pursued without instrumental goals—in part because their main goals are, in themselves, trivial: putting a ball into a net or through a hoop is inconsequential per se. Human societies have, of course, added myriad layers of meaning to certain games, to the point where a ball going into a net can become the most consequential moment of one's life (as when a player scores the winning goal in a soccer World Cup). Yet, in principle, games can be practiced without instrumental goals in mind, something rare for most activities. Take education: despite being intrinsically rewarding, learning is so thoroughly saturated with instrumentality—earning high grades, obtaining numerous credentials, becoming a good career prospect—that pursuing it for its own sake has become structurally difficult and even actively discouraged. Numerous scholars have engaged with normative debates about how enhancing human capacities, which can result in deskilling, can undermine valuable aspects of social activities, including the cultivation of skills, and even become a threat to society or human nature. AI is one such technology, and sport is one such activity. Notwithstanding this, sport seems less affected by the implementation of AI, as complex sporting actions cannot yet be delegated to machines—though robots have recently achieved feats such as racing past human competitors at the Beijing half-marathon or beating elite table tennis players (Gregory 2026; Sample 2026).Footnote 9 Biomedical enhancement, rather than AI, offers a clearer picture of the corrosive effects of deskilling in sport. Sport ethicists provide a wide range of conceptual tools (such as thought experiments and arguments) that help think through the normative implications of biomedical enhancement becoming the main, if not the only, source of human (athletic) performance. For these luminaries, deskilling effects are morally controversial when they undermine the development of valuable human traits associated to the achievement of the meaningful purposes of sport, a position that strongly aligns with arguments against AI use, especially Rüther’s analysis of the meaning gap. I will outline the abovementioned conceptual tools by following Bonte et al. (2014) three “intrinsic objections to doping in sport”: (1) “proper processes”, (2) “proper outcomes”, and (3) “proper origins” (p. 3). These three argumentative strategies propose that the use of performance-enhancing substances or methods eliminates a meaningful human aspect of sport connected to skill exercise: (1) processes intentionally directed by athletes’ own agency allow for meaningful comparisons, (2) effortful strivings produce valuable outcomes in the form of a sense of achievement, and (3) performance resulting from the exercise of innately given capacities leads to meritorious recognition. To these three strategies, I will add a fourth: (4) “proper identity formation”, which hinges on the narrative potential of sport in helping people shape their own, original identities.Footnote 10 2.1 Agency and meaningful comparisons Gleaves (2011) proposes to “imagine a time when a scientist will develop a way for athletes to pharmacologically boost performance without undue health risks” (p. 272). The development of this harm-free, performance-enhancing drug would remove all health-related concerns associated with using such substances to boost athletic performance, which Gleaves considers contingent, and center anti-doping fears on the inherent purpose and value of sport. Drawing on Suits’ views on games and sport, as well as Kretchmar’s distinction between tests and contests, Gleaves highlights the peculiar nature of sport: it specifically presents difficult problems that humans must attempt to solve through the exercise of very concrete skills. One reason sports place such strong onus on skills is epistemological. As Kretchmar (2018) explains, sports can be understood as tests that gauge one’s ability to accomplish difficult tasks. They reveal to athletes what they can and cannot do regarding the exercise of certain physical skills, such as speed, strength, and spatial thinking. In addition to being events for self-understanding, sports can also serve a comparative function when engaged in as contests. Sporting contests allow athletes to determine how well they are doing relative to other performers. Fraleigh (1975) posits that a key purpose of sport is that it “provides non-equivocal knowledge as to the relative status [or abilities] of individuals and of human kind in the world of force, space and time” (p. 80). Similarly, for Loland (2002), “the goal of sport competitions is to measure, compare and rank two or more competitors according to athletic performance” (p. 10). In sum, sporting tests and contests are valuable as epistemological events that allow participants to comprehend how good they are (either individually or comparatively) at exercising a very specific set of abilities. For sport competitions to achieve their epistemological purpose, comparisons must be meaningful; they must pit performances against one another on relevant, measurable, and comparable traits. Put more plainly, they must compare apples with apples and oranges with oranges. To allow for meaningful evaluation, sports have historically sought to create conditions that isolate and measure one variable within the complexity of athletic performance: skill. Meaningful comparisons, then, are grounded in athletic performances that result primarily from skill (improved through training, self-correction, competition, etc.). To this end, equipment or playing conditions are standardized to ensure they equally impact all participants and that the outcome of a contest results mainly, if not exclusively, from the competitors’ exercise of skill rather than from external factors (Loland 2002). Likewise, contestants are classified into categories of similar body type so that no single physical attribute overpowers the rest and the relevant skills play a role (Loland 2021; Parry and Martínková, 2021). Such comparisons are meaningful because they isolate a single relevant, measurable variable, i.e., “naked” skilled athletic performance directed by the athletes’ own agency, while holding everything else constant. From the outlined epistemological vantage point, biomedical enhancement is morally controversial when it undermines the knowledge-based purpose of sport by corrupting measurements, comparisons, and rankings. This is the stance of the United States President’s Council on Bioethics. When evaluating how biomedical enhancement can threaten sport, this group explains that performance enhancement can blur the lines between performance outputs attributable to the athletes’ agency and those attributable to the performance-enhancing substance or method they use. Individuals who do not rely on artificial aids improve their performance through “the self-conscious and self-directed activity that lies at the heart of better training”, that is, through their “own doings” or “self-conscious activity and exertion” (The President’s Council on Bioethics 2003, p. 126). In this case, performance enhancement benefits are clearly attributable to the athletes’ agential capacities. In contrast, when athletes rely on biomedical enhancement that enhances their capacities, they improve “as if by ‘magic,’ without the [abovementioned] self-conscious or self-directed activity” (The President’s Council on Bioethics 2003, p. 126). Artificially enhanced athletes are, at least partially, alienated from their own actions. Their skills “result from having [their] bodies ‘worked on’ by others or altered directly” (The President’s Council on Bioethics 2003, p. 127). Hence, when measuring their performance, it is unclear to what extent their own agency is being assessed.Footnote 11 Biomedical enhancement is problematic when it erodes athlete agency and, in doing so, makes the inherent epistemological purpose of sporting activities unachievable. Other skill-based practices share the same epistemological goal. Education, for instance, can be seen as an activity intended to individually assess and comparatively develop one’s cognitive capacities. Students who use large language models to write their essays produce work that is hardly genuinely theirs, that does not result from their cognitive capacities, undermining the purpose of the educative practice in which they are participating. If a student does not author an essay intended to assess their capacities, the assessment becomes impossible, and both the task itself and the educative practice lose their epistemic value. 2.2 Effortful striving and achievement Suits has his own sport-related thought experiment, arguably the most famous among sport philosophers. He invites his readers to imagine a utopian society (called “Utopia”) in which all instrumental action has been eliminated by the invention of highly efficient, mind-controlled machines, through brain implants, that provide individuals with everything they need and want. In this world, Suits (2005) writes: “Game playing makes it possible to retain enough effort in Utopia to make life worth living … For in games we must have obstacles which we can strive to overcome just so that we can possess the activity as a whole, namely, playing the game” (p. 154, added emphasis). Only games, defined as voluntary attempts to overcome unnecessary obstacles, give individuals an opportunity “to achieve something” when they have nothing to do because all their instrumental needs and wants are instantaneously met (Suits 2005, p. 157, added emphasis). A person who wants a house can simply think of it, and the superefficient machines in Utopia will build it instantaneously. By contrast, Suits refers to John Striver, a Utopian who does not merely want a house; he wants the kind of house he can build himself by exercising his house-building skills. Striver prefers his self-made house because “it would be the kind whose construction would give him the greatest satisfaction, and we may suggest that such satisfaction would require that building the house would provide enough of a challenge to make the task interesting while not being so difficult that John would utterly botch the job” (Suits 2005 p. 156, added emphasis). Suits (2005) equals Striver’s effort to build one’s house to a golfer’s attempt to put balls into holes: “it is clear that John is no more interested in simply having a house than the golfer is in having ball-filled holes. It is the bringing about of these results which is important to John and to the golfer rather than the results themselves” (p. 157, added emphasis). According to Suits, the value of games lies in the process by which individuals try to realize a specific, challenging state of affairs. However, what makes this process valuable within his framework remains contested—in fact, in the direct citations above, I have italicized several sources of value (e.g., effort, difficulty, satisfaction). A plausible interpretation focuses on the notion of difficult achievement. For Kretchmar (2006), Suits’ main lesson is that “we [humans] have been built and rebuilt to solve problems. Thus, we chew on difficulties, we consume tests—and we need to be regularly nourished. When we gaze out onto the world we automatically look for problems and, when we find them, immediately search for points of attack so that we might solve them” (p. 73). Hence, games are sites that provide intrinsically valuable, difficult problems for humans to resolve through skillful performance. Bradford (2015) and Hurka (2014, 2020) have also championed this interpretive line. In his introduction to Suits’ masterpiece, The Grasshopper: Games, Life and Utopia, Hurka locates the value of gameplay in “achievement”, which he defines in terms of difficulty as “accomplishing something in your chosen field”, with the accomplishment involving the process of resolving a complex, challenging problem through skill and ingenuity (Hurka 2014, p. 14). For Hurka (2020), achievement is an objective or perfectionist good whose value is independent of people’s attitudes or subjective states (e.g., deriving satisfaction). This value results from individuals exercising their own powers or skills to “instantiate … a more abstract value of rational, or reason-guided, connection to reality” (Hurka 2020, p. 591). That is to say, achievements are valuable because they produce, through skillful action, an overlap between the mind (i.e., an intended goal) and the world (i.e., a specific state of affairs). Games, in turn, are meaningful as sites for creating the necessary conditions for such achievements. Bradford also explains the objective value of achievement in perfectionist terms.Footnote 12 Though neither Hurka nor Bradford considers the impact of biomedical enhancement on athletic achievement, the implications of their views on this topic are straightforward. Performance-enhancing substances and methods that boost athletes’ capacities can create a misalignment between athletes’ skills and the challenges of their sports. If the obstacles become less challenging, athletes need to exert less effort to overcome them, rendering athletic achievement less valuable. Alternatively, the degree to which mind and world must overlap decreases, reducing the value of sporting accomplishment. Reid (2020), appealing to the spirit of Olympic sport, formulates a perfectionist-based argument against biomedical enhancement. In Olympism, sport is an activity at the service of human development; specifically, sport is “a philosophy of life, exalting and combining in a balanced whole the qualities of body, will and mind. Blending sport with culture and education, Olympism seeks to create a way of life based on the joy of effort” (International Olympic Committee 2026, p. 8). If the joy of effort must drive the type of human development achieved through sport, athletic performance must result from the athletes’ own efforts. If not, the kind of human development that sport promotes is negated. Just as athletic achievement can be diminished when performance-enhancing substances and methods allow athletes to exercise fewer skills or to a lesser degree, AI-powered technologies that enhance human capacities in other areas of life will erode the acquired capacities that lead to valuable achievements in those practices—as well as the role these practices play in serving as sites for meaningful accomplishments. When an agent outsources writing to a language model, the possibility of achievement is negated: the agent neither causes the overlap between mind and world nor exercises the human capacities that Hurka and Bradford identify as valuable—and writing itself, as a practice, ceases to be a site for accomplishments as meaningful as finding the right word, structuring an argument, or inserting a witty comment. 2.3 Natural talent and meritorious recognition Sandel encourages his readers to imagine a robotic, bionic baseball player who benefits from brain implants that perfect the angle and timing of their swing, ensuring that every visit to the strike zone results in a home run. Remarkably, Sandel opens his analysis by dismissing the “agency” approach examined above. In his view, “[t]he problem is not the drift to mechanism but the drive to mastery” (Sandel 2007, p. 27). Put more plainly, for Sandel, the main ethical concern is the erosion of people’s appreciation of their own limitations, not the erosion of technology’s effects on agency. To illustrate his critique of the drive to mastery, Sandel proposes a scenario involving two highly successful and admired athletes. One has earned such recognition through hard work, grit, and determination; the other owes their fame primarily to natural gifts displayed with grace and apparent effortlessness. Typically, people find the latter more admirable than the former. In Sandel’s (2007) words: “No one believes that a mediocre basketball player who works and trains even harder than Michael Jordan deserves greater acclaim or a bigger contract” (p. 29). Sandel then invites readers to consider how recognition would change if both athletes were found to depend heavily on performance enhancement. In this new scenario, the allegedly naturally gifted athlete would be more widely despised than the hard-working one because the “display of natural talents and gifts that are no doing of the athlete who possesses them” is more central to sporting competitions than the exertion of effort (Sandel 2007, p. 28). However, he laments that societies have typically exaggerated the value of striving and effort, depreciating the importance of giftedness. The hubristic overexpansion of human powers has produced a greater appreciation of effort and a corresponding devaluation of giftedness. This overexpansion has grown from the ideas that humans are “self-made” and that their talents and achievements are wholly their own. Biomedical enhancement in sport is a consequence of this anthropological conception. Athletes often go as far as possible to modify their inborn capacities to turn themselves into champions, refusing to accept their inborn limitations. In fact, Sandel points out that, in some athletic circles, failing to embody such behavior is seen as a flaw: “[w]hen performance-enhancing drugs become commonplace, unenhanced ballplayers … are criticized for ‘playing naked’” (Sandel 2007, pp. 88–89). Moreover, athletes whose success comes from their biomedically powered efforts to turn themselves into winners would displace the naturally talented performers that sports are supposed to celebrate. Hence, the wrong kind of competitor would be celebrated, further perpetuating the drive to mastery. Individuals who delegate tasks to AI to accomplish certain goals better than others, failing to accept their own limitations and draw on their inborn capacities, demonstrate the same drive to mastery that Sandel denounces. This hubristic behavior, in Sandel’s (2007) view, is morally problematic because it hinders “key features of our moral landscape—humility, responsibility, and solidarity” (p. 86). When individuals perceive themselves as solely responsible for their talents and achievements, they often develop the conviction that their success is fully their own and entirely deserved. This “triumphalist” attitude toward life leaves no room for “tragedy, mystery, and humility” (Sandel 2020, p. 43). Those who adopt this perspective come to believe that everything that happens to people is their responsibility, not the result of external factors such as natural giftedness or luck. Their victories and defeats are wholly the result of their choices and their mastery over life (or lack thereof for those on the most unfavorable side). In addition to coming at the expense of humility, the overresponsibilization of individuals also undermines solidarity and mutual obligation. In societies that heavily prize individual freedom and achievement, perceiving one’s success or failure solely as one’s own making undermines the sense of solidarity and mutual obligations toward others. Sandel illustrates this point through support and contribution to insurance. Since people are unaware of whether they will be healthy or sick, they pool their risk by buying insurance that can support them if they fall on the unfortunate side and contract an illness. “The result is”, Sandel (2007) points out, “mutuality by inadvertence” (Sandel 2007, p. 90). However, this “solidary” attitude is only sustainable insofar as individuals lack control over their own risks. If individuals become more convinced that they do control such factors, their pursuit of mutualistic practices decreases, in favor of a focus on their own efforts. In Sandel’s (2020) view, the idea of the overly responsible, self-made person is central to today’s liberal societies because they have uncritically embraced meritocracy. Driven by the pursuit of freedom and equality, most societies have openly embraced meritocracy as their preferred method for allocating positions and resources. However, meritocracy favors the ethic of mastery.Footnote 13 Sport, which for some is the social practice that best embodies the meritocratic ideal (Dixon 2021; Fouché 2017, p. 11), epitomizes the ethic of mastery: The meritocratic emphasis on effort and hard work seeks to vindicate the idea that, under the right conditions, we are responsible for our success and thus capable of freedom. It also seeks to vindicate the faith that, if the competition is truly fair, success will align with virtue; those who work hard and play by the rules will earn the rewards they deserve. We want to believe that success, in sports and in life, is something we earn, not something we inherit. (Sandel 2020, p. 125) While appealing, for Sandel, meritocracy rests on faulty assumptions. Individuals can hardly claim that their success (or failure) is entirely their own doing. Sandel (2020, pp. 123–124) illustrates this with sport: LeBron James’ or Usain Bolt’s dominance in their respective athletic pursuits can certainly be attributed only to their efforts; it must also be explained by factors for which they are scarcely responsible, such as being born with natural talents unlike those of other participants. Nor can James’s and Bolt’s advantageous social positions be viewed as self-made. They attracted people’s attention and social recognition because they operated within societies that highly valued their achievements. Had they been born into a society that took little interest in athletics or invested little in sporting activities at which they did not excel, they would hardly enjoy their advantageous social positions. Acknowledging the role these contingent factors (i.e., inborn talent and social preferences) play in their own success would prevent individuals from embracing the ethic of mastery, in favor of an ethic of giftedness, avoiding the negative effects the tyranny of merit has on humanity, responsibility, and solidarity. All meritocratic activities can produce the problems identified above. Without the recognition of the giftedness of life, communities become unable to rightfully appraise and value the talents of their constituents and engage in practices of solidarity that more broadly distribute the results of people’s gifts. Sport, understood as a talent-based practice, teaches societies to accept “the fate of human goods embodied in important social practices … the celebration of natural talents and gifts in athletic and artistic endeavors; humility in the face of privilege, and a willingness to share the fruits of good fortune through institutions of social solidarity” (Sandel 2007, p. 96).Footnote 14 Viewing themselves more humbly as “guardians” or “repositories” of talents that happen to reside in them, instead of owners or authors fully responsible for their existence, individuals realize that they have no special claim on “the fruits of their exercise [becoming more prone] to regard the distribution of natural talents as a common asset and to share in the benefits of this distribution” (Sandel 2008, p. 70). However, the performance-enhancing promise of biomedical enhancement and the pressure toward turning oneself into a champion push against the possibility of viewing sport as Sandel suggests. The same claims apply to social practices in which individuals’ performance relies on AI. The technology’s potential to enhance performance, and perhaps most importantly, to disguise limitations, threatens the acknowledgment of giftedness. If a writer’s talent or a student’s thinking can be improved or even replaced by LLMs, then the (deskilled) practices of writing and studying foster a sense of (fake) mastery, cementing the belief that their (AI-mediated) achievements and resulting fruits are self-made and, therefore, deserved.Footnote 15 Rather than teaching individuals to accept their limitations and mutually work to compensate for them, AI use further promotes the search for individual accomplishment, even if partially (or fully) deskilled and manufactured. 2.4 Narrativity and sport as a meaningful form of life A fourth argumentative strategy centers on the narrative potential of meaningful social activities such as sports. This approach appeals to the contributive value of skilled sporting performance to a central human enterprise: the collaborative shaping of a shared identity. An activity or feature acquires contributive value in virtue of becoming a part of an entity or whole that has some sort of value. Fraleigh (1983) provides the following example: “a soccer ball being kicked into a goal does not have value by itself, but has value as part of a whole called soccer game and has that value because it contributes to the game” (p. 59). Skillful athletic performance has contributive value when its exercise makes the practitioner legible—to themselves and to others—as a genuine participant in a specific form of life: as a cyclist, a writer, a surgeon, or an athlete. The deskilling threat, on this account, concerns the role skillful performance plays in conditions of identity formation and communal intelligibility. In his reflections on the test-contest distinction, Kretchmar contends that the meaning of both sporting tests and contests is predicated on the existence of practice communities. For him, although the presence of the other is clearer in contests (which cannot exist without someone else opposing a performer), “testing is not solipsistic at all” (Kretchmar 2018, p. 232). The intelligibility of a sporting test depends on the existence of testing communities, that is, on the existence of a sport that is overseen and/or promoted by specific institutions (e.g., sport governing bodies, leagues) and engaged in by other individuals who produce collective understandings on how to play the sport, what it means to excel, and why it all matters. For example, a tennis player who achieves a personal best serve speed does so against the backdrop of a practice community that has collectively established what counts as an impressive serve, developed the technical vocabulary to describe and transmit the skill, and instituted the activities (e.g., a tennis match or competition) within which that achievement acquires its meaning. Individuals are attracted to specific tests because they encounter them in these communities, and by engaging in them, they begin to care about their individual testing actions and results (Kretchmar 2000). As such, they begin to shape their identities around common practices and shared meanings, such as consulting certain websites and online spaces linked to their testing community and buying paraphernalia that people in that testing community wear, and frequenting the same spaces (e.g., golf courses, biking trails, and indoor sporting venues) (Kretchmar 2018). In fact, many end up identifying themselves as soccer fans, golfers, bikers, and football enthusiasts. Building on these ruminations on the meaning of sport and incorporating insights from MacIntyre’s notion of social practice, Gleaves (2011) claims that the greatest moral problem with performance enhancement is that it “may change the ‘life of the athlete’” (p. 277), preventing them from becoming or being a genuine member of their sporting social practice. For MacIntyre (2007), social practices have a series of internal goods that can only be achieved by adhering to the standards of excellence and obeying rules established by the practice community and its corresponding institutions. “To enter into a practice is to accept the authority of those standards and the inadequacy of my own performance as judged by them. It is to subject my own attitudes, choices, preferences and tastes to the standards which currently and partially define the practice” (MacIntyre 2007, p. 190). To achieve the internal goods of competitive cycling—the mastery of climbing technique, the strategic reading of a peloton, the cultivated capacity to endure suffering—one must submit to the standards that define the practice: accepting the judgment of coaches and experienced riders, riding within the rules of sanctioned competition (such as anti-doping rules), and subordinating personal preferences to the disciplines of training. Given that today’s sports are governed by anti-doping values and standards that regard (most) biomedical enhancements as wrong, athletes who resort to such means would fail to qualify as genuine participants and would be unable to achieve the internal goods of their sport. As such, they cannot truly identify themselves as members of their testing communities, that is, as golfers, bikers, or table tennis players. Grendreau (2015) posits that those who enhance their performance biomedically are at fault for lying to themselves and, most importantly, to their practice communities and societies about who they are, becoming undeserving of membership in them or of representing (e.g., being featured in) their social practices. For Grendreau, narratives are necessary for individuals to make sense of themselves as individuals and members of collective enterprises. When a person deviates from the shared narratives of their community, they threaten the widely held expectations that define how they recognize one another: As with many social roles, being a sports figure involves having certain components of one’s narrative “ready-made”. While athletes may tell their own stories, those stories are interpreted and understood through the lens of what it means to be an athlete. Included may be a variety of virtues including hard work, being a “team player”, and the general idea of “sportsmanship”. And, independent of whether it is the case for all or even most athletes, what we typically understand as “being an athlete” involves achievement solely through one’s own labor. (Gendreau 2015, p. 524) Today’s sporting communities are dominated by anti-doping expectations. Biomedical enhancement, on this narrativity account, is worrisome because it can become a form of narrative hindrance: it disqualifies enhanced athletes from the story their community tells about what it means to excel. Relying heavily on pharmacological substances or medical interventions delegates to technology work that athletes must accomplish through their own practice. This delegation misrepresents athletes as practitioners they have not become: they acquire a skill-constituted identity they have not in fact cultivated. In doing so, they disrupt the shared intelligibility through which practice communities recognize their members and transmit their standards of excellence. In a similar vein, students who submit language-model-generated essays and programmers who deploy AI-written code raise the same problem: they perform the outputs of a skill-constituted practice without exercising the skill, and in doing so, they erode both their own membership in the relevant community and the community’s capacity to identify who genuinely belongs to it. 3 Conclusion In this article, I have shown that the deskilling debate in AI ethics has precedents in the sport philosophical debate concerning the use of biomedical performance-enhancing means. I have examined four such precedents, each anchored in a distinct source of human value that contributes to the development of meaningful social practices, such as sports. The argument from agency holds that the meaningful exercise of skill requires that performance be genuinely attributable to the individual. If deskilling processes, such as using biomedical enhancement or AI-powered tools, sever the causal link between agency and results, they erode the epistemological purpose inherent to sport—and activities with similar testing and contesting attributes. The argument from achievement insists that competent, difficult effort is the key normative element in rendering sporting achievement valuable. If deskilling eliminates such effort, sporting accomplishments lose value and sport ceases to be a site for meaningful challenges. The argument from natural talent holds that biomedical enhancement displaces appreciation for people’s inborn limitations and talents in sport, resulting in a social practice that contributes to the erosion of humility, responsibility, and especially solidarity by promoting the ethic of mastery. The narrativity argument locates the wrong of deskilling in the disruption of the shared meanings and communal identities through which individuals constitute themselves as practitioners of a form of life, undermining the meaning-making potential of sport. In outlining these four argumentative strategies in sport ethics, which I consider relevant to AI ethics, I have avoided aligning myself with any of them to identify a path forward.Footnote 16 My aim has been more modest: to offer AI ethicists a set of argumentative strategies well-worn in sport ethics (“old plays”) that may prove fresher—and perhaps more generative—in their own (“new”) field. Data availability No datasets were generated or analysed during the current study. Notes Plato lamented the invention of writing for its harmful effects on the capacity for memorization (Ferdman 2025). Similar claims were made regarding the introduction of calculators and the ability to compute mentally, as well as online search systems, such as Google, and the ability to locate and process relevant information (O’Brien 2025; Robbins and Blundell 2025). As one anonymous reviewer noted, I could more broadly focus on technology in sport. An earlier version of this manuscript did adopt that framing. Peer feedback, along with my own reconsideration of my position, convinced me that a narrower focus on biomedical enhancement would better serve my analogical approach. “Technology” is too broad a term: engaging it properly would require examining the nature of technology in sport, categorizing its types, and identifying which types most closely resemble AI use. Biomedical enhancement shares an extremely relevant feature with today’s controversial uses of AI: both are hidden. Some performance-enhancing technologies, such as prosthetic limbs, are visible and easily assessable by third parties; biomedical enhancements are not, since they blend with the body itself. AI use raises the same worry, which, as I will show, runs through every controversial case examined in this manuscript. For an analysis of the different teleological elements of sport, see López Frías (2026). See, for instance, Allen Guttmann’s (1978) analyses of the differences between ancient and modern sports. This assertion is deceiving, however. Philosophers such as Bernard Suits (1982) and Thi Nguyen (2021) have warned against conceptualizing social activities as gamified versions of themselves. Relatedly, the assertion hardly implies that sports are simple; they are extremely complex phenomena that require interpretation from myriad angles. See Graham McFee’s (2004) reflections on sport as a moral laboratory. These machines operated in highly controlled environments that cannot be replicated in many other, more complex sports (e.g., basketball, soccer). Technologies with this potential, such as AI-powered humanoid robot athletes or cyborg athletes, seem more at home in science fiction stories and are still far in the future, should they ever be developed. Even if the latter were the case, the sporting activities they would partake in would hardly be considered human sports, but rather robot sports (López Frías and Luis Pérez Triviño 2016). Hence, AI’s performance-enhancing benefits remain indirect and uncertain in today’s sports (for an examination of these effects, see Carrio 2025; Reed 2022; Zukiewicz 2025). This is why this paper avoids addressing the ethical issues of using AI in sports. Although I regard this as the most promising of the four, space constraints require that I argue for it elsewhere. An insightful comment from one of the journal’s anonymous reviewers prompted me to reflect on the reach of this argument regarding competitions that allow a broader range of biomedical performance-enhancing means, such as the Enhanced Games. The challenge these competitions pose is that giving athletes access to a wider range of biomedical enhancement options would require changing the comparison variable, potentially leading to comparisons of variables that are hardly comparable, for instance, performances resulting from athletes using radically different performance-enhancing means. If, however, universal access and similar effectiveness can be ensured across all competitors, performances might remain comparable and the epistemological purpose of sport preserved (on the difficulties of achieving these conditions, see (Cox and Piatkowski 2024; Holm 2007). Still, this situation would remain problematic for those who claim that sport should be about natural performances. On their view, such competitions would hardly compare human performances, but something else entirely: sport scientists’ proficiency in designing effective biomedical enhancements, very much like the proficiency that Formula One rewards in its car designers and engineers (Ekdahl and Krieger 2024; Hoberman 2026). For her, achievement is valuable because it involves the “excellent exercise of the perfectionist capacities” (Bradford 2015, p. 121), which she defines as those “characteristic of human nature” and “worth developing” (Bradford 2015, p. 116). Specifically, the two capacities she most closely connects to achievement comprise exercising the will and rationality. For example, climbing Mount Everest is a valuable accomplishment primarily because it involves exercising willpower (e.g., overcoming adversity, coping with pain, practicing courage) and rationality (e.g., choosing the best route, knowing when to rest). What makes the state of standing at Earth’s highest peak is the process through which one arrives there. If such a goal were reachable by elevator or helicopter, it would scarcely constitute a valuable achievement—and the activity involving the ascent, which hardly deserves to be called mountain-climbing anymore, would hardly be a site for meaningful accomplishments. Sandel (2020) explains that, throughout the history of Western society, one can see “an uneven contest between an ethic of fortune and a more muscular ethic of mastery. The ethic of fortune appreciates the dimensions of life that exceed human understanding and control. It sees that the cosmos does not necessarily match merit with reward. It leaves room for mystery, tragedy, and humility… The ethic of mastery, by contrast, puts ‘human choice at the center of the spiritual order.’” (p. 43). Sandel acknowledges that celebrating accomplishments that result primarily from inborn capacities is at odds with today’s capitalist societies, especially the United States, where success is supposed to be earned on meritocratic grounds, through one’s hard work, not inherited from external circumstances beyond one’s control. However, this is only the case because individuals have fallen prey to the tyrannies resulting from the misguided assessment of merit prevalent in neoliberal, capitalistic societies (Sandel 2020). This argument can also lead to Brett Karlan’s (2023) worry that the incorporation of AI can exacerbate justice and equity problems by giving those with more resources competitive advantages over less privileged individuals in comparative areas where AI aid can be helpful. Each of these arguments presents justificatory shortcomings. The first argument depends on a contested account of what it means for an action to be genuinely one’s own, a matter that admits of degrees and resists a clean threshold. The second argument must defend the claim that effort or difficulty is intrinsically and objectively valuable. However, this is a perfectionist commitment that many challenge (Dunkle 2025; Hendrickx 2024). Concerning the third argument, Sandel’s argument rests on contested assumptions about the value of giftedness and on conjectural implications of embodying the ethics of mastery. The narrativity argument rests on the controversial assumption that practice communities have stable standards of excellence against which deskilling can be evaluated. Yet, such standards are historically contested and evolving; performance-enhancing means once resisted, from coaching assistance to chemical supplements, were eventually accepted (Morgan 2009). References Aylsworth T, Castro C (2024) Should I use ChatGPT to write my papers? Philos Technol 37(4):117. https://doi.org/10.1007/s13347-024-00809-w Bonte P, Sterckx S, Pennings G (2014) May the blessed man win: a critique of the categorical preference for natural talent over doping as proper origins of athletic ability. J Med Philos 39(4):368–386. https://doi.org/10.1093/jmp/jhu024 Bradford G (2015) Achievement. Oxford University Press, Oxford Carrio A (2025) Artificial intelligence and neuroenhancement in sport, 1st ed. Routledge, London. https://doi.org/10.4324/9781003520290 Cox LTJ, Piatkowski T (2024) Enhancing autonomy through the ‘enhanced games.’ Perform Enhanc Health. https://doi.org/10.1016/j.peh.2024.100296 Dixon N (2021) Sport, meritocracy, and praise. J Philos Sport 48(2):275–292. https://doi.org/10.1080/00948705.2021.1932517 Dunkle ID (2025) Difficulty, effort, and ability. Synthese 206(5):255. https://doi.org/10.1007/s11229-025-05356-1 Ekdahl D, Krieger J (2024) The Enhanced Games: a timely injection for the International Olympic Committee. Front Sports Act Living 6:1490651. https://doi.org/10.3389/fspor.2024.1490651 Ferdman A (2025) AI deskilling is a structural problem. AI Soc. https://doi.org/10.1007/s00146-025-02686-z Fouché R (2017) Game changer: the technoscientific revolution in sports. Johns Hopkins University Press, Baltimore Fraleigh WP (1975) Sport-purpose. J Philos Sport 2(1):74–82. https://doi.org/10.1080/00948705.1975.10654099 Fraleigh WP (1983) An examination of relationships of inherent, intrinsic, instrumental, and contributive values of the good sports contest. J Philos Sport 10(1):52–60. https://doi.org/10.1080/00948705.1983.9714400 Gendreau MS (2015) Who? Moral condemnation, PEDs, and violating the constraints of public narrative. Ethic Theory Moral Pract 18(3):515–528. https://doi.org/10.1007/s10677-014-9536-6 Gleaves J (2011) No harm, no foul? Justifying bans on safe performance-enhancing drugs. Sport Ethics Philos 4(3):269–283 Gregory A (2026) Humanoid robots show rapid advances as they race past humans in Beijing half marathon. The Guardian. https://www.theguardian.com/sport/2026/apr/19/humanoid-robots-race-beijing-half-marathon. Accessed on 22 Aug 2026. Guttmann A (1978) From ritual to record: the nature of modern sports. Columbia University Press, New York Hendrickx M (2024) Analysis of difficulty. PsyArXiv. https://doi.org/10.31234/osf.io/9v8pf Hoberman J (2026) The Enhanced Games: a techno-fantasy that will fail. Perform Enhanc Health 14(1):100380. https://doi.org/10.1016/j.peh.2025.100380 Hofman J, Goldstein DG, Rothschild DM (2023) Steroids, sneakers, coach: the spectrum of human-AI relationships. SSRN Electron J. https://doi.org/10.2139/ssrn.4578180 Holm S (2007) Doping under medical control-conceptually possible but impossible in the world of professional sports? Sport Ethics Philos 1(2):135–145. /z-wcorg/. Retrieved from /z-wcorg/ Hurka T (2014) Introduction. In: Suits B (ed) The grasshopper: games, life and utopia. Broadview Press, Toronto, pp 7–20 Hurka T (2020) The parallel goods of knowledge and achievement. Erkenntnis (Dordr) 85(3):589–608. https://doi.org/10.1007/s10670-020-00245-0 International Olympic Committee (2026) Olympic charter. https://www.olympics.com/ioc/olympic-charter. Accessed on 28 Aug 2026. Karlan B (2023) Human achievement and artificial intelligence. Ethics Inf Technol 25(3):40. https://doi.org/10.1007/s10676-023-09713-x Kretchmar RS (2000) Movement subcultures: sites for meaning. J Phys Educ Recreat Dance 71(5):19–25. https://doi.org/10.1080/07303084.2000.10605140 Kretchmar RS (2006) The intelligibility of suits’s utopia: the view from anthropological philosophy. J Philos Sport 33(1):67–77. https://doi.org/10.1080/00948705.2006.9714691 Kretchmar RS (2018) The nature and value of sporting tests and contests. N Y Law School Law Rev 63:219 Krook J (2025) When autonomy breaks: the hidden existential risk of AI. AI Soc 40(8):6011–6024. https://doi.org/10.1007/s00146-025-02397-5 Lee HP (Hank), Sarkar A, Tankelevitch L, Drosos I, Rintel S, Banks R, Wilson N (2025) The impact of generative AI on critical thinking: self-reported reductions in cognitive effort and confidence effects from a survey of knowledge workers. In: Proceedings of the 2025 CHI conference on human factors in computing systems. ACM, Yokohama, pp 1–22. https://doi.org/10.1145/3706598.3713778 Loland S (2002) Fair play in sport: a moral norm system. Routledge, London Loland S (2021) Classification in sport: a question of fairness. Eur J Sport Sci 21(11):1477–1484. https://doi.org/10.1080/17461391.2021.1923816 López Frías FJ (2025) Games as windows and remedies to modern society: a qualified defense of agonistic encounters. Prax Filos 61:e20214699. https://doi.org/10.25100/pfilosofica.v0i61.14699 López Frías FJ (2026) What goals? Which point? Whose purpose? A critical engagement with sport internalism. Philosophies 11(1):16. https://doi.org/10.3390/philosophies11010016 López Frías FJ, Luis Pérez Triviño J (2016) Will robots ever play sports? Sport Ethics Philos 10(1):67–82. https://doi.org/10.1080/17511321.2016.1166393 MacIntyre AC (2007) After virtue: a study in moral theory, 3rd ed. University of Notre Dame Press, Notre Dame McFee G (2004) Sport, rules, and values: philosophical investigations into the nature of sport. Routledge, London Meier KV (1988) Triad trickery: playing with sport and games. J Philos Sport 15(1):11–30 Morgan WJ (2009) Athletic perfection, performance-enhancing drugs, and the treatment-enhancement distinction. J Philos Sport 36(2):162–181. https://doi.org/10.1080/00948705.2009.9714755 Nguyen CT (2021) How Twitter gamifies communication. In: Lackey J (ed) Applied epistemology, 1st ed. Oxford University Press, Oxford, pp 410–436. https://doi.org/10.1093/oso/9780198833659.001.0001 Nussbaum MC (2013) Creating capabilities: the human development approach. Belknap Press of Harvard University Press, Cambridge O’Brien GD (2025) All play and no work? AI and existential unemployment. J Ethics 29(4):747–771. https://doi.org/10.1007/s10892-025-09522-y Parry J, Martínková I (2021) The logic of categorisation in sport. Eur J Sport Sci 21(11):1485–1491. https://doi.org/10.1080/17461391.2021.1943715 Rafner J, Dellermann D, Hjorth A, Verasztó D, Kampf C, Mackay W, Sherson J (2021) Deskilling, upskilling, and reskilling: a case for hybrid intelligence. Morals Mach 1(2):24–39. https://doi.org/10.5771/2747-5174-2021-2-24 Ratti E, Graves M (2025) A capability approach to AI ethics. Am Philos Q 62(1):1–16. https://doi.org/10.5406/21521123.62.1.01 Reed K (2022) Analytics destroying baseball’s entertainment value. https://troymedia.com/business/analytics-is-destroying-any-entertainment-value-in-baseball/ Reid HL (2017) Why Olympia matters for modern sport. J Philos Sport 44(2):159–173. https://doi.org/10.1080/00948705.2017.1327323 Reid HL (2020) Athlete agency and the spirit of Olympic sport. J Olymp Stud 1(1):22–36. https://doi.org/10.5406/jofolympstud.1.1.0022 Robbins S, Blundell I (2025) Losing our voice? Generative AI and the degradation of human expression. Minds Mach 36(1):2. https://doi.org/10.1007/s11023-025-09757-6 Rüther M (2025) The meaningfulness gap in AI ethics: a guide on how to think through a complex challenge. AI Soc 40(3):1403–1415. https://doi.org/10.1007/s00146-024-01993-1 Sample I (2026) AI-powered robot beats elite table tennis players. The Guardian. https://www.theguardian.com/science/2026/apr/22/ai-powered-robot-beats-elite-table-tennis-players-milestone-robotics. Accessed on 4 Sep 2026. Sandel MJ (2007) The case against perfection: ethics in the age of genetic engineering. Belknap Press of Harvard University Press, Cambridge Sandel MJ (2008) Liberalism and the limits of justice, 2 ed. Cambridge University Press, Cambridge Sandel MJ (2020) The tyranny of merit: what’s become of the common good? Penguin Books, London Suits B (1982) Sticky wickedness: games and morality. Dialog (Can Philos Rev) 21(4):755–759. https://doi.org/10.1017/S0012217300023891 Suits B (1988) Tricky triad: games, play, and sport. J Philos Sport 15(1):1–9. https://doi.org/10.1080/00948705.1988.9714457 Suits B (2005) The grasshopper: games, life and utopia, 2 ed. Broadview Press, Toronto Sung LU, Ferdman A (2025) Empathetic large language models, the social capacities and human flourishing. Inquiry. https://doi.org/10.1080/0020174X.2025.2518448 The President’s Council on Bioethics (2003) Beyond therapy: biotechnology and the pursuit of happiness. Washington, D.C. https://repository.library.georgetown.edu/handle/10822/559341 Vallor S (2015) Moral deskilling and upskilling in a new machine age: reflections on the ambiguous future of character. Philos Technol 28(1):107–124. https://doi.org/10.1007/s13347-014-0156-9 Zukiewicz P (2025) Analytics ruined the NBA. Here’s how they can save it. https://medium.com/@zukiewicz.piotr/analytics-ruined-the-nba-heres-how-they-can-save-it-7f24798d56d9. Accessed on 24 Aug 2026. Acknowledgements I am indebted to the insightful comments from the journal’s reviewer and to the participants at several academic venues where I presented this work for their valuable feedback on earlier iterations of the manuscript. One such venue was the 1st International Congress on the Philosophy of Physical Activity, Sport, and Play, which I attended thanks to funding from the research project “Ética cordial y democracia inclusiva en una sociedad tecnologizada”, reference PID2022-139000OB-C21, funded by the Ministerio de Ciencia, Innovación y Universidades, the Agencia Estatal de Investigación, and the Fondo Europeo de Desarrollo Regional, MCIU/AEI/https://doi.org/10.13039/501100011033/FEDER, EU. Author information Authors and Affiliations Contributions I conducted the research that formed the basis for this manuscript and wrote the full manuscript. Corresponding author Ethics declarations Conflict of interest The authors declare no competing interests AI use All the content in this article is original. None of it has been created using LLMs. This technology has only been used for minor linguistic suggestions, mainly rephrasing or word choice. Additional information Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Rights and permissions Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. About this article Cite this article López Frías, F.J. Old argumentative plays on a new field? Contributions from sport ethics to the ethical analysis of artificial intelligence. AI & Soc (2026). https://doi.org/10.1007/s00146-026-03377-z Received: Accepted: Published: Version of record: DOI: https://doi.org/10.1007/s00146-026-03377-z