What does an institution owe the people whose standing it produces?
Where an institution produces its subjects’ operative standing as entitlement-bearers, its legitimacy depends not on their authorisation but on discharging three obligations they cannot waive: legibility (recognition criteria that fit the material reality of their lives), non-abandonment (not withdrawing the standing it produced without meeting a burden of justification), and remediation (a real path, inside the institution, when recognition fails).
The critique of consent in algorithmic welfare is the diagnosis; the destination is a positive account of what such institutions owe. If a system produces the operative standing of the people it governs, legitimacy cannot be traced back to their authorisation. It has to be located in the institution’s obligations. I call the resulting principle the Constitutive Obligation Principle (COP).
The Constitutive Obligation Principle. Where an institution produces its subjects’ operative standing as entitlement-bearers, its legitimacy depends not on their authorisation but on discharging non-waivable obligations of legibility, non-abandonment and remediation.
Each obligation answers to one of the three structural features of constitutive dependence, rather than being added from outside.
Legibility
Because the institution produces recognition, it cannot treat its criteria of recognition as natural facts beyond its responsibility. The obligation of legibility requires that it make its subjects legible under conditions responsive to the material reality of their lives. A fingerprint criterion that systematically fails the hands of people who do manual labour is a criterion the institution chose, rather than a neutral test that some people happen to fail, and it fails the people the welfare system exists to protect.
Non-abandonment
Because the institution can revise its subjects’ recognition through its own operations, it must not withdraw what it produced without meeting specific burdens of justification. Once it has constituted someone as an operative entitlement-bearer, a cancelled card, a failed link between databases or a lapsed authentication is a withdrawal of recognition rather than a clerical event, and it needs the justification that withdrawing recognition requires.
Remediation
Because the subject has no standpoint outside the institution from which to refuse or contest, she must have a substantive, not merely formal, pathway within it when recognition fails. An institution that answers fingerprint failure by directing the person to an online grievance portal discharges no genuine obligation; it redoubles the failure. The pathway has to reach the person the system cannot currently see.
Why recognition theory
The view locates algorithmic misrecognition within the tradition of Honneth, Fraser and Butler, and recent recognition-theoretic AI ethics, as a distinctively infrastructural form: misrecognition carried not by attitudes or cultural patterns but by the architecture through which recognition is delivered. That is also why I read Aadhaar as an early instance, and not an outlier, of a form towards which algorithmic governance is converging: the obligations it owes are the obligations every such system will owe.
Questions people also ask
What is the Constitutive Obligation Principle?
It is the principle that where an institution produces its subjects’ operative standing as entitlement-bearers, its legitimacy depends not on their authorisation but on discharging non-waivable obligations of legibility, non-abandonment and remediation. I develop it in “Beyond Consent” (AI and Ethics, 2026).
Why can't the people affected waive these obligations?
Waiving is itself an act of authorisation, and the point is that authorisation is not available as a foundation in these cases. An obligation that the dependent party could sign away at the enrolment desk would collapse back into the consent model it replaces.
Are the three obligations a complete list?
They are not offered as one. Each is generated by one of the structural features of constitutive dependence, which is why each is necessary given that feature, but I do not claim that together they exhaust what such an institution owes.
Read next
In the library
Sources and further reading
- Saxena, A. (2026). Beyond consent: Algorithmic welfare, constitutive dependence, and the limits of liberal AI ethics. AI and Ethics 6: 414.doi.org/10.1007/s43681-026-01270-5
- Honneth, A. (1995). The Struggle for Recognition: The Moral Grammar of Social Conflicts. Cambridge: Polity Press.
- Fraser, N. and Honneth, A. (2003). Redistribution or Recognition? A Political-Philosophical Exchange. London: Verso.
- Waelen, R. (2023). The struggle for recognition in the age of facial recognition technology. AI and Ethics 3(1): 215–222.doi.org/10.1007/s43681-022-00146-8
Cite this page
Saxena, A. (2026). What does an institution owe the people whose standing it produces? abhinavsaxena.in. https://abhinavsaxena.in/questions/what-do-institutions-owe/