Figure Ilustration AI
FORMOSA NEWS - Papua - Who Blame When AI Makes Wrong Move New Study Calls For Modernized Machine Liability. The rapid expansion of Autonomous Artificial Intelligence (AAI) systems in high-risk sectors is challenging traditional legal frameworks worldwide, exposing critical gaps in how society attributes responsibility when technology fails. A study published in May 2026 by Yohanna YR Watofa from the Sekolah Tinggi Ilmu Hukum Manokwari in Indonesia reveals that conventional legal doctrines, which depend heavily on proving human fault and clear linear causation, are structurally inadequate to handle the complexities of self-learning algorithms. Because modern AI functions as a "black box" where decision-making processes are obscured, establishing human negligence or tracing a single cause of harm becomes virtually impossible, threatening legal certainty and public safety.
The Growing Crisis of Algorithmic Accountability
As autonomous technologies shift from basic automated software into independent decision-making entities, they are transforming critical sectors such as healthcare diagnostics, algorithmic financial trading, autonomous vehicles, and public administration. Historically, legal systems have relied on fault-based liability and strict liability, both of which operate on the fundamental assumption that a human agent remains in ultimate control. When a human acts with intention or negligence, the law can clearly trace the damage to that individual. Autonomous Artificial Intelligence, however, breaks this traditional link. Modern AI models continuously adapt and learn from new data post-deployment, making their final outputs unpredictable even to the software developers who originally designed them. This shift introduces significant commercial, ethical, and legal complications. If an autonomous vehicle causes an accident or an AI medical system misdiagnoses a patient, the distributed nature of the AI ecosystem which spans software developers, data providers, infrastructure operators, and end-users complicates individual blame, leaving victims vulnerable and the tech industry facing unpredictable legal risks.
Assessing the Legal Divide: A Comparative Approach
The study utilized a normative juridical and comparative legal methodology to evaluate the effectiveness of current accountability doctrines. By analyzing regulatory frameworks across multiple jurisdictions, including the European Union’s Artificial Intelligence Act, regulatory initiatives in the United States, and international guidelines from the OECD and UNESCO, the research assessed how different societies manage autonomous risks. This comparative approach examined classic civil legal codes alongside modern statutes, such as Indonesia's Electronic Information and Transactions Law and its Personal Data Protection Law. The analytical framework evaluated conventional civil structures against emerging regulatory models to identify standard practices for balanced AI governance.
Key Findings: Why Traditional Law Fails
The research demonstrates that traditional liability structures suffer from severe operational limitations when applied to autonomous systems. The breakdown of classical liability involves three core operational pillars:
The Growing Crisis of Algorithmic Accountability
As autonomous technologies shift from basic automated software into independent decision-making entities, they are transforming critical sectors such as healthcare diagnostics, algorithmic financial trading, autonomous vehicles, and public administration. Historically, legal systems have relied on fault-based liability and strict liability, both of which operate on the fundamental assumption that a human agent remains in ultimate control. When a human acts with intention or negligence, the law can clearly trace the damage to that individual. Autonomous Artificial Intelligence, however, breaks this traditional link. Modern AI models continuously adapt and learn from new data post-deployment, making their final outputs unpredictable even to the software developers who originally designed them. This shift introduces significant commercial, ethical, and legal complications. If an autonomous vehicle causes an accident or an AI medical system misdiagnoses a patient, the distributed nature of the AI ecosystem which spans software developers, data providers, infrastructure operators, and end-users complicates individual blame, leaving victims vulnerable and the tech industry facing unpredictable legal risks.
Assessing the Legal Divide: A Comparative Approach
The study utilized a normative juridical and comparative legal methodology to evaluate the effectiveness of current accountability doctrines. By analyzing regulatory frameworks across multiple jurisdictions, including the European Union’s Artificial Intelligence Act, regulatory initiatives in the United States, and international guidelines from the OECD and UNESCO, the research assessed how different societies manage autonomous risks. This comparative approach examined classic civil legal codes alongside modern statutes, such as Indonesia's Electronic Information and Transactions Law and its Personal Data Protection Law. The analytical framework evaluated conventional civil structures against emerging regulatory models to identify standard practices for balanced AI governance.
Key Findings: Why Traditional Law Fails
The research demonstrates that traditional liability structures suffer from severe operational limitations when applied to autonomous systems. The breakdown of classical liability involves three core operational pillars:
- The Epistemic Gap: In fault-based liability, proving negligence requires understanding what an actor knew or should have known. Because machine learning processes are mathematically opaque, even engineers cannot fully explain how a specific automated decision was reached, rendering fault-based proofs impossible.
- Multi-Layered Causation: Traditional law relies on a linear chain of cause and effect. Damage caused by autonomous systems is instead the product of complex, non-linear interactions between training datasets, algorithm designs, and unpredictable operational environments.
- Fragmented Control: Control over an AI system is distributed across an extensive value chain. No single entity retains complete oversight, which makes assigning sole responsibility to an isolated actor disproportionate and unfair.
Transitioning toward a hybrid legal framework that blends risk-based liability with shared responsibility provides clear benefits for policymakers, businesses, and consumers. For industries operating in high-risk sectors like transportation and medicine, a clear, tiered liability model eliminates regulatory ambiguity, allowing developers to innovate safely without fearing sudden, disproportionate litigation. For national governments, including developing digital economies like Indonesia, these findings offer a blueprint for comprehensive legislative updates. Rather than relying on fragmented, sectoral regulations, the study demonstrates the necessity of implementing mandatory algorithmic audits and transparent system testing. This proactive approach ensures that when automated technologies fail, clear channels for legal redress protect human rights and civil safety.
Author Profile
Yohanna YR Watofa holds a degree in law and is a researcher and academic faculty member at the Manokwari Law School in West Papua, Indonesia. Her professional expertise focuses on contemporary legal governance, electronic information laws, and the intersection of civil liability frameworks with emerging autonomous technologies.
Source
Yohana YR Watofa (2026). Human Machine Liability Redistribution in Autonomous Artificial Intelligence Decision Systems within Contemporary Legal Governance. Formosa Journal of Applied Sciences (FJAS) 2026. Vol. 5, No. 5, Halaman 1199-1210
DOI:https://doi.org/10.55927/fjas.v5i5.40
URL: https://journalfjas.my.id/index.php/fjas
DOI:
URL:
0 Komentar