The following essay is being published in conjunction with the “(In)hospitable Space Conference” which explores the human dimensions of space exploration and is cosponsored between the English Department of Carnegie Mellon University, the Society for Social and Conceptual Issues in Astrobiology, and the European Astrobiology Institute.
For most of the space age, the path beyond Earth was narrow. To become an astronaut or cosmonaut, one usually had to pass through national institutions: air forces, space agencies, elite scientific programs, medical boards, and political systems that treated spaceflight as a matter of national prestige. The people selected for flight were not simply passengers or workers. They carried a public meaning. They embodied discipline, courage, technical skill, physical endurance, and the promise that a nation could project its power beyond the atmosphere.
That picture is no longer enough. Spaceflight is still rare, risky, and expensive, but it is no longer organized only around the state astronaut. The new landscape includes private astronauts, commercial passengers, researchers, entrepreneurs, future employees of private stations, and people who may one day apply for work or residence in lunar and Martian environments. Commercial spaceflight participants already differ from traditional NASA astronauts in age, occupation, education, and routes into flight. The social meaning of “spacefarer” is becoming more open.
But openness can be misleading. The fact that more kinds of people may now apply does not mean that access has become fair. Every flight still requires selection, and selection always creates a boundary. There are applications to complete, bodies to examine, psychological profiles to assess, risks to classify, insurance conditions to satisfy, and training programs to enter. Increasingly, these judgments are not made only by doctors, committees, or mission planners. They are also supported by data systems: automated screening tools, risk models, biometric tracking, psychometric instruments, and private scoring systems that claim to identify who is suitable for a mission.
These systems can be useful. Spaceflight is genuinely dangerous. Radiation, launch stress, re-entry, isolation, microgravity, limited medical care, and emergency constraints make some form of screening unavoidable. No ethical argument for wider access can ignore the material reality of spacecraft. Still, danger can become a convenient word. It can hide assumptions that should be examined. A person may be described as too risky when the real issue is that the mission was designed around a narrow idea of the human body. A candidate may be judged a poor “fit” when the score is really rewarding familiarity, social confidence, wealth, or conformity. A decision may look scientific because it is produced by a model, even though the model has inherited the biases of past selection.
This is where the idea of algorithmic hospitality becomes useful. Hospitality is usually associated with welcome, but at a deeper level it concerns control over entry. It asks who is received, who must wait outside, who writes the rules, and who must explain themselves in order to cross a threshold. Philosophers have long treated hospitality as a problem of the stranger, the host, and the border. In spaceflight, that old problem returns in a new form. The “host” may be a space agency, a private company, a medical board, an insurer, or an algorithmic system. The “guest” is the applicant whose body, history, data, and personality are being evaluated.
No spacecraft can be open to everyone. That would be irresponsible. A vehicle or habitat has limited mass, oxygen, space, medical capacity, and emergency options. But necessary limits do not settle the ethical question. The real issue is whether the rules of entry are clear, proportionate, revisable, and open to challenge. Algorithmic hospitality asks whether the gate to space can give reasons. It asks whether exclusion is treated as a technical inevitability or as a decision for which someone remains responsible.
The systems that select people for space are part of the space infrastructure. We often think of infrastructure as rockets, suits, launchpads, radiation shielding, and power supplies. But admission systems also shape the future. They decide which bodies are expected, which risks are tolerated, which forms of difference are accommodated, and which people disappear from the story before the mission even begins.
Medical selection is the clearest example. Astronaut medical standards are necessary, but they are not neutral laws of nature. They depend on the mission, the vehicle, the available medical support, and the level of uncertainty that institutions are willing to accept. A rule appropriate for a long-duration orbital crew may be excessive for a short suborbital flight. A standard needed for a commander may not make sense for a researcher with a limited task. Commercial spaceflight therefore requires more flexible forms of judgment. Too little screening is dangerous; too much screening turns safety into exclusion.
Disability shows why this matters. Traditional astronaut selection often assumed a body that is strong, highly fit, and non-disabled. That assumption is now being questioned. The ESA Parastronaut Feasibility Project has explored how physical performance testing might be reconciled with a more inclusive astronaut corps. Accessibility should be considered throughout mission planning rather than added at the end. This is not about pretending that all bodies face the same risks. They do not. It is about refusing to treat one historical model of the astronaut body as the natural measure of all future participants.
Space itself complicates what ability means. A person limited by terrestrial architecture may function differently in microgravity. Meanwhile, a body considered ordinary on Earth may struggle in orbit. Space is not a natural human environment. It is built, artificial, and hostile. If a built environment excludes someone, the ethical question is not only whether that person is unsuitable. It is also whether the design is too narrow.
Algorithmic tools could support this shift. A well-designed medical model might help replace blanket exclusion with more careful individual assessment. It could show that a managed condition is compatible with a certain mission, or that a person can participate safely with a specific accommodation. But a poorly designed model can freeze the past. If the data come mainly from earlier astronaut populations, the system may learn what previous astronauts looked like rather than what future spacefarers could be. It may mistake historical absence for biological impossibility.
That risk is familiar from debates about algorithmic fairness. Machine-learning systems inherit the world from which their data come. They absorb old categories, missing populations, proxy variables, and institutional habits. Fairness cannot be solved by a single equation, because different ideas of fairness often point in different directions. In spaceflight, the problem is especially sharp because the evidence base is small and shaped by decades of narrow selection.
Psychological screening raises a related concern. Long missions require people who can cooperate, remain calm under pressure, communicate well, and live with others in confined conditions. But the language of “fit” is risky. It may refer to real mission needs, but it can also conceal preference for familiar personalities, emotional styles, or professional backgrounds. A proprietary fit score may appear objective while rewarding people who already know how to sound confident in elite institutions.
For that reason, “fit” should never be left vague. Fit for which task? In which crew? Under which stress? For how long? According to what evidence? If a candidate is excluded because of a psychological or behavioral assessment, the institution should be able to explain the reason in terms that are concrete enough to be understood and, where appropriate, challenged.
This is the heart of algorithmic hospitality: people should not be excluded by systems that cannot explain themselves. Full transparency is not always possible. Some details may involve safety, privacy, or legitimate commercial protection. But opacity cannot be the default. A person rejected by an automated or semi-automated process should have access to a meaningful explanation. They should be able to correct errors. They should know whether an appeal or reassessment is possible. And institutions should track where people are being filtered out.
This last point is crucial. Exclusion does not happen only at the final “yes” or “no.” It can occur much earlier: in recruitment language, application costs, educational requirements, medical forms, training access, physical tests, visa rules, insurance conditions, or interviews that reward certain social styles. A person may never apply because the invitation already signals that they do not belong. A fair system must therefore map its own exclusions.
The rise of biomedical data makes the issue even more delicate. Projects such as the Space Omics and Medical Atlas promise a deeper understanding of how human bodies respond to spaceflight. This research may improve care, monitoring, and countermeasures for future crews. Yet it also raises a question that will become more urgent as spaceflight expands: how much of oneself must one reveal to be allowed to go? Data needed for safety should not be quietly mixed with data collected for research, insurance, marketing, or future model training. Consent becomes weak if refusing broad data use means losing access altogether.
Public–private spaceflight makes accountability harder still. A company may own the selection platform. A public agency may license the mission. A medical board may approve the candidate. An insurer may shape the risk threshold. If someone is unfairly rejected, responsibility can scatter. Each actor may point to another. Algorithmic hospitality resists that scattering. It insists that selection systems used in publicly licensed, partnered, or meaningful missions should include documentation, explanation, appeal, and clear responsibility.
The stakes become even higher when the issue is not a single mission but a habitat. Future stations, lunar bases, and Mars settlements will need admission rules. Early habitats will need specialists, and people who can live with limited resources. But habitats will also decide whose bodies are planned for, whose needs are normal, and whose risks are accepted.
Selection, then, is not just a filter. It is a founding act. If the first habitats are built around narrow assumptions, later inclusion will look costly or impractical. If early crews are selected mainly through wealth, elite education, military experience, or cultural familiarity, those traits may later appear to be natural requirements. If proprietary algorithms help decide who belongs, inequality may be built into off-world life before anyone names it as inequality.
Space exploration often speaks of new beginnings. But a new place does not guarantee a new politics. Old exclusions can travel in medicine, design, law, capital, and data. The task is not to remove every boundary; spaceflight will always need boundaries. The task is to make those boundaries answerable. Who gets to go is not a question that comes after the engineering is complete. It is one of the ways engineering becomes ethics. If human beings are to build a future beyond Earth, the gate to that future must be treated as part of the habitat itself.
