LONDON (Realist English). The global refined products trading market is undergoing a deep structural transformation. Unlike the mature crude oil market, trading in diesel and gasoline in major consumer regions such as Asia has long been controlled by a few producers, refineries, shipping companies, and large commodity traders, while information remained relatively opaque. However, the rapid adoption of artificial intelligence technologies is breaking down this structure.
AI Tools Come to the Forefront of Trading
Geneva-based commodity analytics firm Sparta Commodities recently introduced a decision-making tool called “Leonidas AI,” capable of offering specific trading and freight recommendations for international markets in response to policy changes within seconds.
During one demonstration, Leonidas AI assessed the potential consequences of a US ban on diesel exports and generated five trading recommendations, sorted by confidence level, covering several markets, including one freight recommendation — the entire process took just a few seconds.
Sparta CEO Felipe Elink Schuurman stated that the new winners will be those trading desks that convert data into trading actions fastest.
In parallel, companies such as Kpler and Windward, engaged in shipping tracking and data, are also adopting AI, turning data into actionable strategies or analyzing the movements of suspicious vessels, enabling new investors, including hedge funds, to overcome information barriers.
Response of Traditional Traders
Traditional traders are also accelerating their transformation. Shipergy, a marine fuel trading company, is deploying AI tools to avoid increasing staff during a market downturn.
CEO Daniel Rose reported that the company has reorganized commercial operations around three hubs — Singapore, Athens, and New York — making Athens the main operational base, while using automation to handle more business without increasing personnel costs.
Rose noted that the compliant addressable market has shrunk significantly, and sanctions, especially against Russia, have reduced available trading volumes. “Over the past three years, we have built the core of fuel trading — trade processing, liquidity, credit compliance, and other infrastructure work well. This core can handle far more business than now, but I don’t want to increase staff and costs for that.”
Practice of Korean Companies
South Korean LPG import giants SK Gas and E1 are also introducing AI into their overseas trading operations. SK Gas’s Singapore trading team numbers only about 5 people but brings the company huge profits annually. The company is developing an AI program that, by studying the past profitable models of experienced traders and global supply and demand data, provides traders with an “AI advisor.”
E1 has also launched an AI-driven transformation, hiring AI engineers to improve its working systems. The company’s overseas trading organization numbers about 10 people and has a branch in Singapore.
In the first quarter of this year, E1’s overseas LPG trading business showed losses, as traders accumulated inventories in the Middle East, and the blockade of the Strait of Hormuz disrupted major trade flows — traders’ decisions directly influenced the divergence in the overseas results of SK Gas and E1.
Reshaping the Competitive Environment
The intervention of AI is reshaping the competitive environment in fuel trading. The traditional advantage lay in logistics capabilities — owning vessels and storage, controlling cargo flows. But in the new environment, these capabilities no longer guarantee success.
The decisive advantage lies in how quickly a company can interpret data, allocate capital, and execute decisions.
This transformation is also changing the pricing mechanism. Commodity markets are no longer driven solely by physical supply and demand, but reflect a complex interaction of algorithmic positioning, real-time data analysis, and rapid capital flows. Information is absorbed instantly, almost without delay. But speed also introduces noise — signals can be amplified, and volatility can spike.
Challenges of Trust and Transparency
The widespread use of AI in fuel trading faces a key challenge of trust and transparency. Industry analysts note that the success of AI systems will depend on explainability and transparency.
Traders must be able to challenge forecasts, verify assumptions, and inject human context into machine outputs. Systems that work as “colleagues” rather than black boxes can inspire the trust needed for real adoption.
Some analysts warn that AI recommendations may become crowded or ineffective, and old traders may resist new entrants. At the Asia-Pacific Petroleum Conference, executives discussed the growing role of AI in data analysis, shipping, and inventory management, but also acknowledged that human relationships and judgment remain indispensable.
Assessment
Fuel trading is not in decline but in restructuring. Logistics, finance, and data are becoming inseparable. The winners will be those capable of operating in three dimensions — combining the reach of physical assets, the precision of algorithms, and the power of capital.
For traditional traders, the main risk is inaction. Just as financial traders and hedge funds are already using the speed and adaptability of AI, physical participants who wait too long may lose their competitive advantage. The winning strategy: start small, test quickly, scale what works — because in a market where milliseconds win, the future belongs to those who learn faster.
