Legal Liability of Autonomous Artificial Intelligence Systems in Emerging Regulatory Frameworks A Comparative Perspective from Indonesia

Figure Ilustration AI

FORMOSA NEWS - Lampung Timur - Indonesia Faces Legal Gaps in Autonomous AI Liability Compared to European Union Standards, Study Finds. Indonesia's current legal framework lacks specific provisions to determine legal responsibility when autonomous Artificial Intelligence (AI) systems cause financial, social, or physical harm. A comparative legal study published in May 2026 by researchers Yanto Irianto from Universitas Nahdlatul Ulama Cirebon, Joko Suranto from STIS Darul Ulum Lampung Timur, and Lulu Sarini from the Kementerian Agama Provinsi Sulawesi Tengah evaluates Indonesia's regulatory readiness against the European Union Artificial Intelligence Act. The findings highlight that Indonesian law remains fragmented across sectoral regulations, creating a legal responsibility gap for autonomous algorithmic decisions. Establishing clear AI liability rules is essential to protect consumers, foster public trust, and provide legal certainty for tech companies operating in emerging digital economies.

Background: The Need for Autonomous AI Liability Frameworks

Autonomous AI systems are expanding across high-risk sectors worldwide, including financial technology, autonomous vehicles, healthcare diagnostics, biometric surveillance, and public administration. Unlike traditional software, autonomous AI operates independently without direct human control, continuously learning and making automated decisions. When automated algorithms cause economic loss, biased outcomes, privacy breaches, or systemic harm, traditional legal systems struggle to assign legal responsibility among software developers, platform operators, data providers, and end users. While international jurisdictions like the European Union have implemented proactive, risk-based frameworks to balance innovation with consumer protection, developing nations like Indonesia face significant regulatory uncertainty. Addressing this regulatory gap is vital for digital economic growth, consumer safety, and human rights protection.

Methodology: Comparative Normative Juridical Analysis
The researchers conducted qualitative legal research using a normative juridical method, incorporating statutory, conceptual, and comparative approaches.

• Legal Materials Analyzed: Primary legal materials included Law Number 11 of 2008 on Electronic Information and Transactions (ITE Law), Law Number 27 of 2022 on Personal Data Protection (PDP Law), Law Number 8 of 1999 on Consumer Protection, the European Union Artificial Intelligence Act, and the General Data Protection Regulation (GDPR).
  
• Analytical Approach: The authors systematically classified and compared regulatory provisions regarding AI risk classifications, algorithmic accountability mechanisms, strict liability models, and transparency requirements.
  
• Verification and Credibility: Source validation and document triangulation were applied across statutory texts, international AI governance guidelines (OECD and UNESCO), and peer-reviewed cyber law literature.
  

Key Research Findings

• Absence of Autonomous AI Liability Rules: Indonesian law applies human-centered legal responsibility models that cannot effectively address self-learning AI algorithms, creating a clear responsibility gap when automated decisions cause harm.
  
• Fragmented Sectoral Framework: Existing Indonesian statutes—such as the ITE Law and PDP Law—focus on general electronic system security and personal data processing. They lack specific provisions for AI risk classification, automated decision-making liability, or conformity assessments.
  
• Opaque Algorithmic Accountability: Most AI implementations in Indonesia operate as "black box" systems, lacking legal mandates for algorithmic transparency, mandatory audits, or explainability rights for affected individuals.
  
• Weak Digital Consumer Protection: Digital consumers in Indonesia remain in a disadvantageous position relative to tech companies controlling algorithms. The current Consumer Protection Law does not establish duties for service providers to explain automated decision-making logic.
  
• Contrast with European Union Governance: The EU AI Act establishes a structured, risk-based classification model (minimal, limited, high, and unacceptable risk) with mandatory human oversight and preventive compliance obligations under Article 9, Article 13, and Article 14.
  

Real-World Impact and Policy Implications
Establishing an integrated, risk-based AI legal framework benefits both society and industry. For technology developers and operators, explicit liability assignment removes legal ambiguity, encouraging safe digital innovation and institutional investment. For consumers, mandatory algorithmic transparency ensures fair treatment in credit scoring, employment recruitment, digital healthcare, and e-commerce platforms. Policymakers and legislators can utilize these findings to draft AI-specific regulations, integrate explainability rights into digital governance, and establish cross-sectoral oversight bodies capable of managing algorithmic risks in emerging markets.

Researcher Profiles
Yanto Irianto, S.H., M.H., is a legal scholar and faculty researcher at Universitas Nahdlatul Ulama Cirebon, specializing in technology governance, cyber law, and comparative legal studies.
Joko Suranto, S.H., M.H., is a legal researcher and academic at STIS Darul Ulum Lampung Timur, specializing in digital law, regulatory policy, and administrative legal systems.
Lulu Sarini, S.H., M.H., is a legal specialist at the Kementerian Agama Provinsi Sulawesi Tengah, focusing on public policy analysis, digital regulation, and legal compliance.

Source
Yanto Irianto, Joko Suranto,Lulu Sarini. Legal Liability of Autonomous Artificial Intelligence Systems in Emerging Regulatory Frameworks: A Comparative Perspective from Indonesia. International Journal of Law Analytics (IJLA). Vol. 4, No. 2, hlm. 349-368
DOI: https://doi.org/10.59890/ijla.v4i2.224
URL: https://journal.multitechpublisher.com/index.php/ijla/index