Aligning Institutional Power and Operational Ability

Andrea Marino; Giovanni Sileno; Thomas van Binsbergen; Tom van Engers

cs.AIOct 27, 2026

Aligning Institutional Power and Operational Ability

Authors: Andrea Marino, Giovanni Sileno, Thomas van Binsbergen, Tom van Engers

Organizations: Informatics Institute, University of Amsterdam, The Netherlands

Abstract

To be socially sustainable, Agentic AI systems require normative designs that align responsibilities with the ability to fulfil them. Identifying incongruences such as unsatisfiable expectations or privilege escalation is crucial in avoiding normative pitfalls. In this study, we discuss design dimensions suitable for describing multi-agent systems in light of this norm-ability alignment requirement. We propose a formal framework centred on process-oriented modelling and institutional power for the analysis of a given normatively relevant process, discuss strategies for repairing it, and elaborate on how the symmetrical relation between operational process and institutional state affects the lifecycle of normative positions.

Figures & tables

trigger	activate	deactivate
$\bot$	$\mathsf{pow}(O,\mathsf{instruct})$ , $\mathsf{pow}(B,\mathsf{sell})$ , $\mathsf{pow}(C,\mathsf{bankrupt})$ , $\mathsf{pow}(O,\mathsf{terminate})$	–
$\mathsf{instruct}$	$\mathsf{duty}(B,\mathsf{sell})$	$\mathsf{pow}(O,\mathsf{instruct})$
$\mathsf{sell}$	–	$\mathsf{duty}(B,\mathsf{sell})$
$\mathsf{terminate}$	–	$\mathsf{pow}(B,\mathsf{sell})$ , $\mathsf{pow}(O,\mathsf{terminate})$
$\mathsf{bankrupt}$	$\mathsf{pow}(C,\mathsf{seize})$	$\mathsf{pow}(B,\mathsf{sell})$ , $\mathsf{pow}(C,\mathsf{bankrupt})$
$\mathsf{seize}$	–	$\mathsf{pow}(C,\mathsf{seize})$

Table 1: The policy

\Pi

with initial facts (trigger

\bot

) and other reactive rules. Most powers associated to owner

O

, broker

B

, and creditor

C

are one-off ;

\mathsf{pow}(B,\mathsf{sell})

is standing and is revoked only by

\mathsf{terminate}

\mathsf{bankrupt}

Panel of Figure 1 — Figure 1: Operational net $N$ and governed reachability graph of the consignment example. The marked places ( $\bullet$ ) in figure 1(a) give $M_{0}$ . The dashed midline divides the broker’s sale process (top) from the creditor’s seizure process (bottom), with the contested goods straddling both. The symbol $\leftrightarrow$ abbreviates an equal-weight input/output pair, representing a read test. A subset of the institutional state required to showcase incongruent scenarios based on the policy of Tab. 1 is presented in Fig. 1(b) ; $\varnothing$ marks none of the shown positions active. Each node pairs the marking $M$ with the institutional state $I$ ; edges are fired transitions. The coloured (and labelled) nodes are classified against the alignment property introduced in section 4 .

Panel of Figure 2 — Figure 2: Repaired net and partial governed reachability graph. Amended arcs are dashed and blue. The transitions sell , terminate , and bankrupt consume active ; instruct reads it. With the amended policy, every reachable configuration is aligned. Only the initial instruct branch and selected state components are shown.

AI Agent Governance

AI agent governance studies how to authorize, constrain, monitor, and hold autonomous and multi-agent systems accountable.

See all 442 papers

Anchoring Distributed Agency: Institutional Metaphors for Designing Large-Scale AI and AI Techniques for Running Institutions
Mar 9, 2018Alexander Boer, Giovanni SilenoAI Agent Architecture Multi-Agent Systems
KPMG, Amsterdam, the Netherlands · Complex Cyber Infrastructure, Informatics Institute, University of Amsterdam, Amsterdam, the Netherlands
Agency and Architectural Limits: Why Optimization-Based Systems Cannot Be Norm-Responsive
Feb 26, 2026Radha SarmaAI Governance AI Alignment
Social, Legal, Ethical, Empathetic and Cultural Norm Operationalisation for AI Agents
Mar 12, 2026Radu Calinescu, Ana Cavalcanti, Marsha Chechik +2AI Alignment Responsible AI
University of York, York, UK · University of Toronto, Toronto, Canada · Polytechnique Montréal, Montréal, Canada

Mar 9, 2018J/K move · Enter open · S save

Anchoring Distributed Agency: Institutional Metaphors for Designing Large-Scale AI and AI Techniques for Running Institutions

Alexander Boer, Giovanni Sileno

KPMG, Amsterdam, the Netherlands · Complex Cyber Infrastructure, Informatics Institute, University of Amsterdam, Amsterdam, the Netherlands

AI Agent Architecture Multi-Agent Systems Agent-Based Modeling

Open paper AskPDF arXiv

Artificial Intelligence (AI) started out with the ambition to reproduce the human mind, but as soon as the sheer scale of that ambition became apparent, it retreated into either studying specialized intelligent behaviours or proposing overarching architectural concepts for interfacing specialized intelligent components---conceived of as agents acting through organizational structures. The agent-based modelling paradigm proved to have interesting applications beyond the original domain, supporting the understanding, simulation, and the assessment of behaviour of socio-economic and legal structures on an aggregate level. Looking at contemporary AI, whereas text completion provides a successful example of specialized behaviour, surprisingly little of the rich foundation of multi-agent systems (MAS) and normative systems research is currently reflected in Agentic AI systems. Addressing this gap, the chapter examines cross-cutting conceptualisations, modelling problems, and design challenges shared by large-scale distributed AI and institutional systems. Three main contributions can be highlighted. First, by tracing the historical trajectory of AI, the chapter identifies an ecological perspective on intelligence, examining both the value and limits of the compression-as-intelligence hypothesis. Second, it shows how moving between internal cognitive mechanisms and external manifestations of agency supports a functional separation of concerns for both individual and collective agency. Third, it argues that epistemological, normative, and value pluralisms are necessary conditions for sustaining the ecological character of intelligence. Furthermore, we indicate how AI techniques can support institutions by modelling social roles, simulating normative interactions, detecting and diagnosing failures, and supporting procedures of review and revision.

Abstract

AI Agent Governance

Explore similar work

Anchoring Distributed Agency: Institutional Metaphors for Designing Large-Scale AI and AI Techniques for Running Institutions

Agency and Architectural Limits: Why Optimization-Based Systems Cannot Be Norm-Responsive

Social, Legal, Ethical, Empathetic and Cultural Norm Operationalisation for AI Agents