This Article argues that legitimacy is an autonomous regulatory objective, distinct from alignment and not secured by it, which seats consequential AI rule-setting in venues a polity already treats as authoritative.
Abstract
AI systems already govern. They rank speech and allocate attention, filter applicants and triage claims. The dominant frame for AI governance, alignment, asks whether such systems pursue the right objectives safely. It cannot answer a prior question: by what right are those objectives set and enforced? This Article argues that legitimacy is an autonomous regulatory objective, distinct from alignment and not secured by it. Legitimacy here is sociological: the belief among those subject to power that it is exercised rightfully. Performance does not produce that belief. We already have the proof of concept. Social media and search delivered enormous gains on every familiar metric and still triggered a legitimacy crisis, because publics questioned who authorized a handful of firms to set the rules of speech, visibility, and knowledge. It is possible to build a benevolent AI and still face a political crisis over its authority. The Article maps three sites where AI legitimacy falters: opacity, which blocks audiences from forming justified beliefs; private power, where firms exercise public-facing authority without recognizable authorization; and administrative automation, which strains reason-giving, participation, and review inside the state. It then asks what law can contribute. Thin legality (publicity, stability, consistent application) signals non-arbitrariness and buys real recognition, but invites legitimacy-washing when form drifts from practice. Thick legality supplies what form cannot: public authorship of the rules that bind. Three portable principles follow. Integration seats consequential AI rule-setting in venues a polity already treats as authoritative. Familiarity presents rules and reasons in locally credible forms. Contestation guarantees a credible second look with real remedies.
The institutional use of artificial intelligence (AI) in parliamentary work is expanding from transcription, translation and document management to research assistance, amendment analysis and public input processing. Yet the accountability mechanisms governing such use remain underdeveloped. Existing responsible-AI frameworks emphasise transparency, human oversight and risk management, but they are largely designed for public administration, market actors or generic public-sector deployment and only partially address parliaments as autonomous legislative institutions. This article asks who governs AI use within parliaments and where democratic accountability gaps arise when algorithmic systems mediate legislative work. Using a conceptual-normative approach grounded in theories of democratic legitimacy, political accountability and algorithmic governance, the article argues that the central problem is not algorithmic autonomy but the displacement of politically relevant choices into infrastructural systems that shape information, options and procedural conditions before formal legislative decisions are made. The article develops a four-part typology: epistemic opacity, institutional attribution, procedural oversight and democratic-legal contestation gaps.
Artificial intelligence is increasingly promoted as a tool for modernizing public administration, accelerating decision-making, improving public services, and reducing administrative costs. Yet, in heterogeneous Global South contexts shaped by structural inequality, technological dependency, unequal access to digital infrastructure, and uneven institutional capacity, algorithmic efficiency may also generate new forms of democratic exclusion. This article develops a normative conceptual analysis of AI governance and argues that public uses of AI should not be evaluated primarily through technical efficiency, ethical compliance, or procedural safeguards, but through democratic legitimacy. It proposes the concept of democratic algorithmic legitimacy, understood as a relational property of the sociotechnical and institutional arrangements through which public authority is exercised with the support of AI. Such arrangements are legitimate when their purposes and operation can be publicly justified to affected persons, when those persons have meaningful opportunities to influence and contest their use, and when responsible institutions retain the authority and capacity to review decisions, repair unjustified harms, modify systems, suspend their operation, or withdraw them when necessary. The framework operationalizes this standard through seven interdependent dimensions: transparency, participation, inclusion, accountability, contestability, correctability, and social justice. This conceptual architecture distinguishes technical performance from democratic authority and explains why efficient outcomes cannot compensate automatically for exclusion, opacity, weak accountability, inaccessible contestation, or ineffective correction. The article identifies interconnected structural, institutional, social, and democratic risks associated with AI deployment in unequal sociotechnical environments and outlines a governance agenda based on meaningful public participation, democratic impact assessment, independent scrutiny, institutional guarantees of explanation, review and appeal, protection of affected groups, public control, technological capacity, and context-sensitive regulation. The article concludes that AI governance should be assessed not only by what computational systems optimize, but by whether societies retain the democratic authority to shape, question, supervise, correct, and, when necessary, reject their use.
A. Duche-Pérez, Marco Tulio Falconí Picardo, Emmanuel Neptalí Augusto Chávez Urquizo et al.· Frontiers in Political Scien...· 0 citations
AI governance instruments are proliferating, and so are their difficulties. Across major jurisdictions and international bodies, reform efforts built on substantially different premises encounter a recognizably similar pattern of failure. I argue that anticipatory regulatory governance rests on three
operational premises
—categorical stability, epistemic accessibility, and manageable pace—and that AI’s emergence, opacity, and velocity violate all three in compound. These premises form a distinct layer of operational preconditions, not a complete theory of governance. Reform within the existing premises reproduces the violations they produce. I call this configuration the
reform trap
: a paradigmatic lock-in at the level of operational preconditions, distinct from path dependence and policy paradigm rigidity. The pattern is convergent across five strategies in active reform—categorical regulation, process-based management, information disclosure, normative guidance, and adaptive experimentation—and persists even in the most adaptive of them. I propose three
premise-level substitutions
—outcome observability, causal attributability, and enforcement capability—each replacing a violated precondition with a weaker, design-addressable one. These differ from outcome-based and performance-based regulation, which swaps instruments within an architecture whose premises remain stable.
It is argued that for advanced AI systems deployed in high-stakes environments the more urgent question may be prudential and strategic, and there is a threshold of evidential and strategic risk beyond which it becomes rationally justified to adopt norms of treatment that include constraints on coercion, deletion, and instrumental use.
The question of how to govern well has been a central topic of inquiry in both political science and philosophy. We contribute to this discourse by examining the epistemic foundations of good governance within democratic and autocratic systems. Drawing on the existing debate surrounding the epistemic justification of democratic procedures (Estlund, 2000; Goodin & Spiekermann, 2018; Landemore, 2012), we construct a thought experiment where agents are tasked with collectively identifying some optimal collective decision. We hypothesize that, for such tasks, democratic decision making systems possess an institutionally driven epistemic advantage compared to autocratic ones. To explore this hypothesis, we develop a simulation where a group of agents sets out to estimate the optimal level of public good provision. To this end, group members engage in individual learning, collective deliberation and voting. We then assess and compare two distinct mechanisms for aggregating individual estimates to the group level: one reflecting a democratic process and the other an autocratic setting. Our findings reveal three key insights. First, democratic decision-making outperforms autocratic methods in terms of judgment accuracy. Second, democratic decision-making processes turn out to fare best when individual citizens are not impartial but instead employ a mild bias towards their own needs when assessing the optimal level of public good supply. A last finding is that, under certain conditions, restricting the duration of deliberation can enhance the epistemic precision of collective decisions.
Dominik Klein, Johannes Marx· Synthese· 0 citations
In the debate on artificial intelligence, the term “AI winter” is often used as shorthand for a technical failure or a temporary decline in market interest. This article proposes a different interpretation. We argue that the classical AI winters were primarily legitimacy crises, in which the system of justifications linking technical promise, funding, commercialization, and social acceptability collapsed. Drawing on the history of AI, legitimacy studies, and the sociology of expectations, we propose a hierarchical interaction model of four legitimacy gaps. In this model, the capability gap functions as the technical substrate of AI promises; the institutional assessment and commercialization gaps mediate whether those promises are credited, funded, and productized; and the governance gap operates as a meta-condition of legal, moral, and political authorization. On this basis, we reinterpret the first and second AI winters. We argue that the episodes later grouped under the first AI winter can be read primarily as a crisis of capability and assessment, whereas the second was a crisis of product, brand, and commercial ecosystem. We then argue that the contemporary boom of generative models does not herald a simple repeat of past winters. A more likely scenario is a regulatory-economic cooling driven by compliance costs, disputes over training data, infrastructure concentration, information manipulation, algorithmic verification, and increasing documentation requirements. In this scenario, generative AI becomes contested not only as an automation technology, but also as epistemic infrastructure involved in producing, verifying, ranking, and stabilizing public truth. In the final section, we formulate four implications for governance: promise restraint, evidence of deployment, documentation obligations, and greater infrastructural pluralization. This perspective shifts the debate from the question of whether AI works to the question of under what conditions its development remains politically and ethically legitimate.
Mariusz Mazurek, Jacek Gurczyński· AI and Ethics· 0 citations
