AI-based technological transformation as a driver for development of oil refining market: case study of Indonesia

This study investigates the multifaceted relationship between AI-driven technological transformation and the demand for downstream petroleum products in achieving Indonesia’s long- term economic growth goals, aligning with the «Golden Indonesia 2045» vision. Employing a mixed- methods approach, the research quantitatively assesses the immediate impact of AI on downstream petroleum operational efficiency (the first hypothesis) and its subsequent influence on critical macroeconomic indicators like GDP growth and the oil and gas trade balance (the second hypothesis). Concurrently, it qualitatively examines the strategic alignment of national AI policies, such as the National AI Strategy from 2020 to 2045 (Strategi Nasional Kecerdasan Artifisial, Stranas KA) and the «Making Indonesia 4.0» roadmap, with downstream energy development plans (the third hypothesis), while identifying associated implementation challenges. Findings reveal a significant positive correlation between AI adoption and improved operational efficiency within the downstream sector (supporting the first hypothesis). This is substantiated by evidence of sophisticated AI applications, including predictive maintenance (PdM) powered by advanced computational methods, which ensures continuous operation and extends the life of critical hydrocarbon assets. Furthermore, AI-integrated fuel blending systems demonstrate high precision, achieving a coefficient of determination (R2) of 0.99 during validation, which showcases robust real-time optimization capability that surpasses traditional modeling and reduces waste. However, the analysis of macroeconomic leverage provides only partial support for the second hypothesis. While AI-influenced efficiency — by maximizing domestic output and optimizing costs — shows a statistically significant, albeit moderate, positive impact on reducing the oil and gas trade deficit and boosting GDP growth, this effect is severely limited by persistent structural issues. Specifically, petroleum imports have a large and negative impact on Indonesia’s economic growth. The operational savings are currently dwarfed by the volume of necessary imports and the enormous fiscal burden imposed by incomplete fuel subsidy reforms, which peaked at 2.8% of GDP in 2022. The oil and gas trade balance persists in a deficit, recording -1.55 billion USD in May 2025 and -1.58 billion USD in July 2025, even amidst an overall national trade surplus. The study confirms a strong top-down strategic alignment between national AI and energy sector policies. Nevertheless, significant implementation hurdles highlight the necessity for targeted policy intervention (supporting the third hypothesis). These pervasive barriers include chronic infrastructure gaps, weak data governance frameworks, severe digital skills shortages requiring systematic improvement from foundational education, high initial investment costs and profound organizational inertia within large enterprises, leading to a «pilot trap», where successful small-scale projects fail to scale up due to cultural and systems integration difficulties. Ultimately, these challenges temper the transformative potential of AI, shifting its current role primarily towards improving operational efficiency within the legacy system. For AI to become a driver of fundamental structural change — the necessary process of reallocating labor and resources toward higher-productivity modern industries — policy interventions must link AI investment to comprehensive energy subsidy reform and the acceleration of the new and renewable energy sector. This research bridges a critical gap in the literature by offering an integrated analysis of technology adoption in a resource-dependent emerging economy, providing evidence-based recommendations for policymakers and industry leaders to effectively leverage AI for sustainable and structural economic growth.