The sudden outbreak of conflict involving Iran has accelerated the integration of artificial intelligence (AI) tools into financial markets. Leveraging their high-speed information processing and scenario analysis capabilities, these systems are evolving from supplementary aids into essential decision-making partners for traders overwhelmed by vast data streams. In the initial days following the escalation, Maxence Visseau, founder of investment firm Arkevium, fully incorporated AI into his investment decision-making process to assess the market impact. By utilizing large language models, Visseau reduced his research time by approximately 80%. He employed Anthropic's Claude to conduct multiple scenario stress tests simultaneously, compare historical precedents, and map out potential ripple effects across various asset classes. Visseau, a Dubai-based macro trading strategist, stated, "I worked nearly 48 hours straight, monitoring interceptions in the UAE while modeling various market scenarios to prepare for the market open. It's precisely in these critical moments that AI becomes indispensable." While Visseau acknowledged that AI cannot yet reliably replace human judgment, he emphasized that the efficiency gains are increasingly vital amid market turmoil triggered by a conflict that has disrupted global energy supplies and resulted in at least four thousand fatalities.
Interviews with investors and strategists globally indicate this conflict has further embedded AI tools into trading workflows, even as many point out potential issues like imprecise prompting and resultant inaccuracies. "We are witnessing history—the first major conflict where AI is being used in warfare while traders rely on it in unprecedented ways to model the war's contours," said Nick Twidale, Chief Market Analyst at AT Global Markets in Sydney, who has 25 years of trading experience. The most significant advantage of using AI tools like OpenAI's ChatGPT, Google's Gemini, and China's DeepSeek is the dramatic optimization of time management. Jian Shi Cortesi, a portfolio manager at GAM Investment Management in Zurich, noted that whereas she might previously spend half an hour reading information from various sources to follow news, she can now get a summary of the latest war developments in seconds. Gathering information on a specific company, which once took days, can now be completed in a day or less. "Doing research before was like digging a hole with a shovel; now it's like using a large excavator. Efficiency has increased at least fivefold," Cortesi described.
Another major advantage is the ability to mine historical data almost in real-time, providing context and reference points for potential market movements, which is especially crucial in the current volatile environment. Following repeated attacks on key Middle Eastern energy infrastructure and heightened risks of escalation, Brent crude surged by up to 11% on Thursday, surpassing $119 per barrel. In Sydney, Anna Wu, a cross-asset strategist at Van Eck Associates Corp., used ChatGPT and Claude to analyze a century of historical data, identifying past oil price surges triggered by wars and determining the best-performing asset classes during different phases. To enhance the analysis, she had the AI cross-reference data with median inflation and global economic growth figures. "AI has significantly boosted efficiency. Previously, historical analysis meant sifting through countless Google search results; now, the time required is drastically reduced," Wu said.
For Gustavo Pessoa, founding partner of São Paulo hedge fund Legacy Capital Gestora de Recursos Ltda., AI tools provide immediate access to information that was previously difficult to gather, which is increasingly critical for assessing the investment implications of a potential de facto blockade of the Strait of Hormuz. "We are using AI for everything, from analyzing ship types and calculating oil demand elasticity to estimating the number of barrels needed to stabilize crude flows," Pessoa stated. However, AI is not infallible and cannot replace human experience and decision-making. Errors have occurred in applications ranging from game development to news content presentation. A policymaker at the Bank of England warned that widespread use of AI in trading could amplify market volatility and herd behavior. Visseau of Arkevium stressed the importance of continuously verifying the accuracy of AI outputs. "It's an iterative process. I question the AI's conclusions, refine the assumptions, and introduce new data dimensions," he explained.
Michael Brown, Senior Research Strategist at Pepperstone Group in London, believes that deep subject matter expertise is crucial for leveraging AI effectively, adding that the technology is not a "magic bullet." "Market participants themselves still need a profound understanding of the situation to make final trading decisions and to discern when an AI model might output false information, which happens periodically," Brown said. Nevertheless, the improving efficacy of AI tools suggests that roles like junior research analysts may face existential threats. Cortesi admitted she no longer relies on junior analysts, calling AI her "most capable" research assistant. "I can ask AI to outline a company's key points using Warren Buffett's investment logic, and it's done instantly. If I assigned that to a junior employee, they might not even understand what Buffett's logic entails. For complex requests like this, AI far outperforms humans, and it's much faster," she noted.
Whether AI will remain solely an assistive tool in the long term remains uncertain. Even before the Iran conflict, investor concerns about AI's displacement effects triggered sell-offs in stocks across sectors, from software firms to food delivery platforms. John Foo, founder of Singapore's Valverde Investment Partners, stated that, for now at least, AI is far from replicating human thought patterns. "AI is an additional tool that assists human research; it's a complement, not a replacement. Good judgment and experience are still needed in the decision-making process, something AI cannot comprehend at this stage, nor likely in the next two to three years. What happens beyond that is anyone's guess," Foo concluded.
Comments