From Trading Floors to Robo-Advisors, AI Is Transforming How Securities Markets Work

The IMF finds that artificial intelligence is rapidly becoming central to securities markets, improving efficiency, lowering costs, and widening access, but also increasing risks related to volatility, opacity, data concentration, and harm to retail investors. It calls for proportionate, capacity-aware regulation that strengthens supervision, transparency, and international cooperation so innovation does not undermine market integrity or financial stability.

The IMF Technical Note Regulatory Considerations Regarding Accelerated Use of AI in Securities Markets draws on work by the International Monetary Fund, the International Organization of Securities Commissions, the Financial Stability Board, the CFA Institute, and research bodies such as the Alan Turing Institute to explain how artificial intelligence is rapidly becoming embedded in modern capital markets. The paper stresses that AI is not entirely new to finance, but recent advances in machine learning and generative AI have dramatically increased the speed, scale, and complexity of automated decision-making. Cheaper computing power, cloud services, and massive data availability have pushed AI from experimental tools into core market infrastructure, forcing regulators to confront risks that older supervisory frameworks were never designed to handle.

Where AI Is Being Used in Capital Markets

AI is now present across the full investment cycle. In asset management, firms mainly use AI to analyze large and messy data sets, identify patterns, and support research rather than to fully automate investment decisions. Alternative data, such as news, filings, and social media, is increasingly used to gauge market sentiment. In wholesale trading, algorithmic and high-frequency trading already account for a large share of transactions, improving speed and efficiency but also increasing sensitivity to shocks. Robo-advisory services use algorithms to profile clients, build portfolios, and rebalance investments at low cost, making financial advice more accessible. Neo-brokers combine digital trading, personalization, and low fees to attract retail investors, while crowdfunding platforms use AI to match investors with projects, predict campaign success, and automate communication. Together, these uses show that AI is reshaping not just trading but also how investors enter and interact with markets.

Benefits That Come with Hidden Risks

The paper is clear that AI can improve efficiency, lower costs, and widen access to financial services. Faster trading, better risk analysis, and cheaper advice can support market development, especially in emerging markets and developing economies. However, these benefits come with important risks. AI systems can amplify market volatility because they react quickly and often in similar ways. Many models are difficult to understand or explain, making it hard for firms and regulators to predict behavior in stressed conditions. Heavy reliance on data raises concerns about quality, bias, and privacy, while the use of similar data and models across firms can lead to herd behavior and synchronized trading that worsens market swings.

Challenges for Regulators and Supervisors

Regulators face a growing gap between fast-moving technology and slower-moving rules. Monitoring AI-driven markets requires skills in data science, model validation, and market microstructure that many authorities, especially in EMDEs, lack. Data itself is often fragmented, expensive, or incomplete, limiting effective oversight. The paper highlights concentration risks, where large firms with access to superior data and computing power gain persistent advantages, creating uneven competition. A particularly difficult issue is unintentional coordination: AI systems can “learn” strategies that look like collusion without any explicit agreement, falling into a legal gray area. Retail investors are also more exposed to harm, as AI enables highly targeted advice, marketing, and even fraud that can exploit behavioral biases and low financial literacy.

How Authorities Can Respond Going Forward

Rather than calling for heavy-handed regulation, the IMF recommends a practical and proportionate approach. Supervisors should build basic AI skills, use simple and scalable monitoring tools, and focus on the biggest risks rather than trying to track everything in real time. Transparency is key: firms should clearly explain how AI is used, what its limits are, and where humans remain in control, especially for retail-facing services. Authorities are encouraged to strengthen cooperation across borders, share information on risks, and pay close attention to concentration and third-party dependencies in AI infrastructure. The central message is that AI will continue to shape securities markets, and regulation must evolve with it to protect investors and market stability while still allowing innovation to deliver its promised gains.