AI in the Stock Market: Hype or Hidden Risk?

By Donggeon Kim (Grade 11)

With the latest developments in technology, artificial intelligence has become the central piece to many modern-day financial markets. Financial firms nowadays use these AI systems to increase internal efficiency, with tasks such as predictive analytics and high-speed algorithms having been optimized. 

Furthermore, many of these firms are utilizing AI based on the fear of being left behind in the competition, rather than relying on data that actually suggests these systems can outperform their human counterparts. At the same time, rising demand for AI systems suggests a speculative bubble is forming, one that has the potential to collapse if AI proves to be ineffective later down the line. These are the dangers that AI poses to the finance industry today. 

As stated by the Bank for International Settlements, this newly formed speculative bubble exists largely due to supply and demand-side factors that are created from AI incorporation within firms. On the supply side, improvements in large language models and machine learning have allowed companies to filter through vast amounts of data much faster than any human employee ever could.

Meanwhile, on the demand side, firms use AI to reduce costs by automating mundane tasks, giving them an advantage within financial markets. In addition, the Financial Stability Board (FSB) emphasizes that financial firms have used AI to optimize portfolios, improve regulatory compliance, and increase operational speed; however, the rapid pace of AI integration, specifically the use of generative AI and large language models, means that firms are investing their resources before establishing relability in AI, which could hold future risks as reliance increases. Furthermore, a 2025 journal article highlights similar concerns, stating how, despite a strong ability to discern historical patterns, AI models tend to fail in real-time markets, where changes in investor behavior and macroeconomic shocks can render such historical patterns obsolete. This journal article additionally notes how these newer AI models are especially susceptible to overfitting, a condition where AI has been trained to follow past data too closely, failing to fully capture the underlying patterns and consequently leading to AI performing poorly when analyzing live markets. 

Moreover, the consequences of this bubble formation can be seen during the 2010 “Flash Crash”, where a large sell order triggered a chain reaction in automated trading algorithms that were programmed to react to price changes and large orders. When this sell order hit the market, these algorithms started to rapidly sell stocks, creating a destructive feedback loop that caused the Dow Jones Industrial Average to drop nearly 1,000 points within minutes and wiped out roughly a trillion dollars in value across American industries. This event highlights the vulnerability of automated markets when firms rely too heavily on algorithms that interact unpredictably. With modern AI now responsible for faster execution and with less human discretion, the consequences of such market shocks in the future could be exacerbated. 

In the present day, the FSB warns that signs of an AI bubble may already be forming, evinced by rising investments in AI infrastructure, inflated valuations of AI companies, and an increasing concentration of AI vendors that create a dependency for many firms. In addition, The Guardian reports that many central bankers have echoed these concerns, warning that inflated valuations could trigger market destabilization if investor sentiment shifts. 

There is, therefore, a serious risk in the future if no caution is taken when integrating AI into finance. Despite the fact that AI utilization in firms can facilitate improvements in insights, speed, and efficiency, much investment is based on hope rather than proven results, leading to potential consequences.

The danger of AI development lies not in the utilization of AI itself, but in the overconfidence held by these firms, where opaque, unproven models are being trusted without sufficient oversight. This is why, as AI continues to develop, we should treat the implementation of such technologies with humility, transparency, and risk management to ensure stability within the future financial markets.