AI is at war and also "commanding" transactions: a Middle East conflict has disrupted Wall Street's workflow.

Facing the sudden Iran conflict, artificial intelligence (AI) tools, with their efficient information processing and situational analysis capabilities, are gradually transforming from auxiliary tools into indispensable decision-making partners for traders overwhelmed by massive information. In the initial days after the outbreak of the Iran conflict, Maxence Visseau, founder of investment firm Arkevium, integrated AI comprehensively into his investment decision-making process to understand the impact of the conflict on the market. Using large language models, Visseau reduced his research time by about 80%. He used Anthropic's Claude to conduct stress tests on various scenarios, compare historical precedents, and identify potential chain reactions that could affect various asset classes. Visseau, who is based in Dubai and focuses on macro trading strategies, said, "I stayed up for nearly 48 hours, closely monitoring the UAE's interception operations while simulating various market scenarios to prepare for the market opening. It was at these critical moments that the role of AI became indispensable." Although Visseau admitted that AI cannot reliably replace human judgment, he believes that in the context of a conflict disrupting global energy supply and causing at least four thousand deaths that has led to violent market fluctuations, the efficiency improvement brought by AI is increasingly indispensable. Interviews with investors and strategists from around the world show that this conflict has further integrated AI tools deeply into trading workflows, although many have pointed out potential problems such as inaccurate prompts and distorted results. "We are witnessing history the first AI being used for warfare, and traders relying on AI in unprecedented ways to analyze the war situation," said Nick Twidale, Chief Market Analyst at Sydney's AT Global Markets, with 25 years of trading experience. Using AI tools like ChatGPT from OpenAI, Gemini from Alphabet Inc. Class C (GOOGL.US), and DeepSeek from China, the most significant advantage is the significant optimization of time management efficiency. Jian Shi Cortesi, a fund manager at GAM Investment Management in Zurich, said that she used to spend half an hour reading information from different sources to keep up with the news, but now, she can get a summary of the latest war developments in just a few seconds. Gathering information on a specific company now only takes a day or less, compared to several days in the past. "Research used to be like digging a hole with a shovel, but now it's like operating a large excavator, with efficiency increased by at least five times," described Cortesi. Another advantage is the ability to almost real-time explore historical data, providing references and background for future market trends, which is crucial in the current environment of volatile markets. Due to multiple attacks on key energy facilities in the Middle East and escalating conflict risks, Brent crude oil skyrocketed by 11% on Thursday, exceeding $119 per barrel. Deep tracing of historical data Anna Wu, a cross-asset strategist at Sydney's Van Eck Associates Corp., used ChatGPT and Claude to trace back a hundred years of history, identify oil price spikes caused by past wars, and determine the best-performing asset classes in each phase. To enhance the analysis value, she also had AI cross-validate data such as the inflation median and global economic growth rate. "AI has indeed greatly improved efficiency; in the past, historical analysis required searching through Alphabet Inc. Class C's results, but now it takes much less time," said Wu. For Gustavo Pessoa, founding partner of So Paulo's hedge fund Legacy Capital Gestora de Recursos Ltda., the use of AI tools is crucial in evaluating the impact of the Hormuz Strait blockade on investments, providing him with information that was previously difficult to obtain. "From analyzing ship types, calculating oil price demand elasticity, to estimating the number of barrels needed for stable crude oil circulation, we are using AI in every aspect," Pessoa said. However, AI is not perfect and cannot replace human experience and decision-making. From game development to news presentation, AI has made mistakes. A policy maker at the Bank of England warned that widespread use of AI in trading could amplify market volatility and herd behavior. Visseau from Arkevium emphasized the need to continuously verify the accuracy of AI output results. "It's an iterative process; I question the conclusions given by AI, adjust assumptions, and introduce new data dimensions," he said. Entry-level analyst positions facing challenges Michael Brown, Senior Research Strategy at London's Pepperstone Group, believes that to fully utilize AI, a strong foundation of expertise is essential, and added that this technology is not a "cure-all." "Market participants still need to understand the situation deeply to make final trading decisions and discern whether the AI model is providing false information, which happens occasionally," Brown said. However, as the efficiency of AI tools continues to improve, entry-level research analyst positions may face survival challenges. Cortesi admitted that she no longer relies on junior analysts, stating that AI is her "most reliable" research assistant. "I can have AI summarize the core points of a company using Warren Buffett's investment logic immediately; but if I assign this to junior employees, they may not even understand what Warren Buffett's investment logic is. For complex demands, AI performs much better than humans and much faster," she said. Whether AI will remain only as an auxiliary tool in the long run remains uncertain. Before the outbreak of the Iran conflict, investors were concerned that the disruptive effect of AI could impact various industries, leading to sell-offs in stocks from software companies to food delivery platforms. John Foo, founder of Singapore's Valverde Investment Partners, said that at least for now, AI is far from replicating human thought patterns. "AI is an additional tool for human research, a supplement rather than a replacement. Because good judgment and experience are still necessary in the decision-making process, something that AI cannot understand at this stage and even in the next two to three years. As for how it will develop afterward, nobody can predict."