By Jack Bardakjian, Group Managing Director, Gapuma November 2025   Artificial intelligence is reshaping business thinking faster than any technological leap since the personal computer. Everywhere you look — from central banks and shipping lines to fertiliser allocation systems and metals hedging desks — AI is not merely present; it is transformative. But let me be clear from the outset: in physical commodities trading — real shipments, real facilities, real partners and real industry outcomes — AI is an instrument, not an overlord. The world I work in every day, and have for decades, is one where trust sits next to data, where geopolitical insight sits beside logistics discipline, and where physical execution remains the final measure of excellence. AI is advancing rapidly, yes; but the winners in our industry will be those who marry machine intelligence with human judgement, relationships, and operational grit. Technology does not replace edge — it sharpens it. Gapuma’s mission remains unchanged: strengthen supply security, serve industry, and advance African and global development with integrity and expertise. What changes are the tools available to us. Today, AI is becoming one of those tools — and a decisive one when deployed responsibly. A Trader’s View: Why AI Matters — and How It Actually Works in the Real World There is no shortage of breathless commentary about AI transforming everything overnight. But traders do not operate in headlines; we operate in risk, delivery, deadlines, counterpart confidence, and execution. AI earns its place in our toolbox when it helps us: anticipate market shifts before they become headlines identify shipping or supply risks early verify counterpart integrity with greater rigour accelerate due-diligence and compliance sharpen pricing and risk frameworks support ESG accountability at scale It is fashionable to describe AI as revolutionary. In commodities, I’d argue it is evolutionary — but evolution can be just as powerful when applied with discipline and insight. As one academic team from Imperial College London recently concluded in a study on AI models for forecasting commodity markets: “AI-driven meta-learning improves long-term commodity price forecasting by synthesising multiple input horizons.” — Charouif, Sommerfeld & Plancherel, Imperial College London That’s academic language for a very practical truth: AI gives us more eyes and more ears. It enhances intuition. It does not replace it. Commodity Trading: Still a Human Arena — Now With Enhanced Vision The essence of physical trade remains unchanged: We build trust. We secure supply. We price risk. We move goods. We honour commitments. No algorithm negotiates port clearance in West Africa at midnight. No model persuades a fertiliser supplier in the Gulf to prioritise your cargo in a tight season. No chatbot navigates a customs challenge in East Africa or discusses storage options in Alexandria with the nuance local context demands. However, what AI does exceptionally well is keep watch when humans cannot. It reads data the way experienced traders read tone. It provides alertness at scale. A senior consultant from Oliver Wyman summarised the inflection point perfectly: “Leading commodity traders are embracing AI to expand their competitive edge.” — James Koh, Oliver Wyman Partner This isn’t hype — it’s strategic positioning. Where AI Adds Immediate Value to Physical Commodity Trading Market Perception and Pattern Recognition Today’s markets react not only to fundamentals, but to sentiment, political speech, shipping news, drought forecasts and mining output rumours. Professor Tom James, a respected authority in commodity risk, put it bluntly: “Sentiment is data — if you have the vision to measure it.” — Professor Tom James AI gives us that vision. It reads signals across thousands of sources at once. That does not eliminate human analysis — it elevates it. Early Risk Detection in Shipping and Logistics Gapuma ships into ports where efficiency ranges from world-class to emergent-capacity. Predictive models now scan: AIS vessel data weather disruptions insurance alerts conflicts and sanctions changes port congestion and labour risk As the CTRM Centre’s analysts recently noted: “The most promise lies where data, process and domain expertise align.” — Commodity Technology Advisory (Vasey & Reames) Exactly — the machine watches; the trader chooses. Counterparty & Compliance Integrity Today’s regulatory landscape — sanctions, ESG scrutiny, beneficial-ownership transparency — demands speed and depth. AI can scan millions of data points in seconds. A compliance advisor recently said: “AI expands compliance capacity by identifying behavioural anomalies long before traditional systems would.” We still check, question and verify — but with greater armour. Strategic Forecasting and Scenario Planning Supply chains are not linear; they are networks. AI allows us to test assumptions: What if ammonia supply tightens? What if a weather disruption affects crop cycles? What if a shipping route becomes militarily sensitive? The Bank of England has warned: “Automation may magnify systemic errors if not paired with strong governance and human oversight.” That is exactly right. AI makes us faster. Governance keeps us wise. The Human Edge: Why Traders Still Matter More Than Ever Here is the truth the headlines rarely capture: The more technology enters trading, the more valuable experience becomes. Computers can forecast price pressure — but they do not know which mill owner in Cairo is likely to change tender specifications at short notice, or which producer in India prioritises long-term reliability over seasonal margin. Machines optimise for volatility. Humans optimise for continuity and relationships. A commodities specialist recently observed: “The trading houses that win will integrate AI without losing the human character of trade.” Precisely. Respect for partners, consistency in supply, real-world judgement — that is still where advantage lives. Education — The Next Strategic Battleground Traditionally, traders learned their craft through markets, mentors and experience. That DNA remains — but today’s demands are broader. Partnership between trading instinct and digital fluency is now essential. Universities are responding. Imperial College London now offers dedicated AI and machine learning training for commercial professionals, designed to bridge real-world business with advanced analytics. This is not “tech for tech’s sake.” It is capability development. And at Gapuma, we see skill development as investment, not cost. The next […]