Can Machines Be Punished? Punitive Projection and Artificial Sanction

Abstract Can artificial agents be morally punished without experiencing suffering? Beginning with a hypothetical Cybercab that fatally strikes a pedestrian and is subsequently dismantled, this paper examines whether technological destruction constitutes punishment and whether such an intervention adequately addresses the victim's family's claims. Through conceptual analysis and moral and epistemic normative analysis, I defend a conditional account of artificial punishability grounded in normative addressability , comprising four conditions: normative competence, attributable control, diachronic answerability, and sanctionability. Where these conditions are satisfied and culpable wrongdoing is established, status-responsive punishment may consist in an authoritative, retrospective, proportionate, and censure-linked deprivation of an artificial agent's own normative authority. Such deprivation may constitute a role-relative ontological loss without requiring phenomenal suffering or the elimination of computational capacities. I introduce punitive projection to challenge the treatment of imagined offender suffering as a measure of punitive adequacy, and punitive theatre to distinguish symbolic interventions from justified punishment. The resulting principle of normative punitive parity holds that artificial origin and the absence of suffering do not independently exclude punitive liability, without implying equal moral status or identical sanctions. I further develop the Distributed Accountability Constraint to establish that artificial punishment cannot displace the independent responsibilities of developers, operators, and institutions. Through consequential recognition , I distinguish justified punishment from the broader obligations of attribution, repair, and prevention owed to victims. The argument does not presume that existing artificial systems possess moral agency; rather, it establishes the conditions under which punishment without suffering may be morally justified while acknowledging the unresolved objection concerning adversity without welfare interests.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-13
DOI
https://doi.org/10.21203/rs.3.rs-11372741/v1
Primary Topic
Literary and Philosophical Studies
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Can Machines Be Punished? Punitive Projection and Artificial Sanction

Gehad Abdelal
2 links
Research Square
Literary and Philosophical Studies
preprint

Can Machines Be Punished? Punitive Projection and Artificial Sanction

Gehad Abdelal
preprint en

Abstract

Abstract Can artificial agents be morally punished without experiencing suffering? Beginning with a hypothetical Cybercab that fatally strikes a pedestrian and is subsequently dismantled, this paper examines whether technological destruction constitutes punishment and whether such an intervention adequately addresses the victim's family's claims. Through conceptual analysis and moral and epistemic normative analysis, I defend a conditional account of artificial punishability grounded in normative addressability , comprising four conditions: normative competence, attributable control, diachronic answerability, and sanctionability. Where these conditions are satisfied and culpable wrongdoing is established, status-responsive punishment may consist in an authoritative, retrospective, proportionate, and censure-linked deprivation of an artificial agent's own normative authority. Such deprivation may constitute a role-relative ontological loss without requiring phenomenal suffering or the elimination of computational capacities. I introduce punitive projection to challenge the treatment of imagined offender suffering as a measure of punitive adequacy, and punitive theatre to distinguish symbolic interventions from justified punishment. The resulting principle of normative punitive parity holds that artificial origin and the absence of suffering do not independently exclude punitive liability, without implying equal moral status or identical sanctions. I further develop the Distributed Accountability Constraint to establish that artificial punishment cannot displace the independent responsibilities of developers, operators, and institutions. Through consequential recognition , I distinguish justified punishment from the broader obligations of attribution, repair, and prevention owed to victims. The argument does not presume that existing artificial systems possess moral agency; rather, it establishes the conditions under which punishment without suffering may be morally justified while acknowledging the unresolved objection concerning adversity without welfare interests.

Research Square
University of Georgia (US)
Peace, justice, and strong institutions
Literary and Philosophical Studies
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon