Attributing Criminal Liability for Harm Caused by Semi-Autonomous Artificial Intelligence: Subjectlessness and International Legal Benchmarks

This article re-examines the attribution of criminal liability for harm caused by semi-autonomous artificial intelligence (AI) in light of international legal principles governing human rights protection and the regulation of AI-related activities. It defines subjectlessness as a criminal-law situation in which a socially harmful outcome is directly associated with an algorithmically generated decision, yet cannot immediately be attributed to a single identifiable person without compromising the principles of culpability, personal liability, and the prohibition of liability without fault. Functional autonomy, data processing, adaptation, and independent selection of means do not establish criminal-law personhood. The study examines the limits of conventional legal concepts of the offender, fault, and causation, while distinguishing individual criminal liability from states’ obligations to protect human rights, ensure accountability, and provide effective remedies. A developer, manufacturer, supplier, deployer, owner, operator, or user cannot be held criminally liable solely because of a formal connection with a system. The article proposes a multilevel functional attribution model that identifies the lifecycle stage at which an impermissible risk arose, the person competent to address that risk, their knowledge and control, a culpable breach of duty, and its causal significance. The proposed directions for regulatory development include differentiated lifecycle duties, risk-based requirements, legally specified meaningful human control, narrowly defined offences addressing particularly dangerous breaches involving high-risk systems, and requirements for algorithmic traceability. The analysis supports a functional, anthropocentric approach that neither personifies algorithms nor permits human actors to invoke algorithmic autonomy to conceal culpable human decisions. Its consistency with international human rights safeguards requires effective protection for affected persons while preserving the substantive and procedural conditions of individual criminal liability.

Authors

Publication Details

Journal
Archives des sciences a multidisciplinary journal
Published
2026-11-20
DOI
https://doi.org/10.68304/as/76515
Primary Topic
Digitalization, Law, and Regulation
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Attributing Criminal Liability for Harm Caused by Semi-Autonomous Artificial Intelligence: Subjectlessness and International Legal Benchmarks

Олег Карась, Yuliia Fidria, Vitalij Grabovets, Olha Hanych et al.
Archives des sciences a multidisciplinary journal
Digitalization, Law, and Regulation
article

Attributing Criminal Liability for Harm Caused by Semi-Autonomous Artificial Intelligence: Subjectlessness and International Legal Benchmarks

Олег Карась, Yuliia Fidria, Vitalij Grabovets, Olha Hanych, Nataliia Kozak
article en

Abstract

This article re-examines the attribution of criminal liability for harm caused by semi-autonomous artificial intelligence (AI) in light of international legal principles governing human rights protection and the regulation of AI-related activities. It defines subjectlessness as a criminal-law situation in which a socially harmful outcome is directly associated with an algorithmically generated decision, yet cannot immediately be attributed to a single identifiable person without compromising the principles of culpability, personal liability, and the prohibition of liability without fault. Functional autonomy, data processing, adaptation, and independent selection of means do not establish criminal-law personhood. The study examines the limits of conventional legal concepts of the offender, fault, and causation, while distinguishing individual criminal liability from states’ obligations to protect human rights, ensure accountability, and provide effective remedies. A developer, manufacturer, supplier, deployer, owner, operator, or user cannot be held criminally liable solely because of a formal connection with a system. The article proposes a multilevel functional attribution model that identifies the lifecycle stage at which an impermissible risk arose, the person competent to address that risk, their knowledge and control, a culpable breach of duty, and its causal significance. The proposed directions for regulatory development include differentiated lifecycle duties, risk-based requirements, legally specified meaningful human control, narrowly defined offences addressing particularly dangerous breaches involving high-risk systems, and requirements for algorithmic traceability. The analysis supports a functional, anthropocentric approach that neither personifies algorithms nor permits human actors to invoke algorithmic autonomy to conceal culpable human decisions. Its consistency with international human rights safeguards requires effective protection for affected persons while preserving the substantive and procedural conditions of individual criminal liability.

Archives des sciences a multidisciplinary journalVol. 76(5)
Digitalization, Law, and Regulation
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
Attributing Criminal Liability for Harm Caused by Semi-Autonomous Artificial Intelligence: Subjectlessness and International Legal Benchmarks — Олег Карась, Yuliia Fidria, et al. · Archives des sciences a multidisciplinary journal (2026) | TGRS Research Map | TGRS