Artificial intelligence in algorithmic trading: A simple theoretical model and market structure implications

• AI reshapes trading via faster data processing and improved prediction models. • AI enhances market efficiency but may introduce new risks and temporary distortions. • This paper presents a theoretical model linking AI strategies to market microstructure outcomes. • This paper also includes empirical illustration with LSTM forecasting to show practical impact. • This paper contributes to understanding AI's effects on financial stability and regulation. Artificial intelligence (AI) is reshaping algorithmic trading by accelerating information processing, enhancing predictive modeling, and altering core elements of market microstructure. Recent advances demonstrate that machine learning systems can refine order-flow interpretation and short-horizon return forecasting, thereby influencing liquidity formation, volatility dynamics, and the speed of price discovery (Ansari, 2024; Kearns & Nevmyvaka, 2013). At the same time, policy analyses highlight that AI-driven trading may introduce new operational and systemic risks, raising important questions for market stability and regulatory design (U.S. Department of the Treasury, 2024; The Financial Stability Board, 2024; Adrian, 2024). This paper develops a simple theoretical model linking AI-based trading strategies to informational efficiency under realistic frictions, supported by clarified assumptions, data requirements, and validation procedures. A brief empirical illustration using an LSTM forecasting setup demonstrates how model-driven signals integrate into execution rules and shape microstructure outcomes (Pilla & Mekonen, 2025). Taken together, the framework shows that AI narrows certain inefficiencies while potentially creating new, temporary distortions arising from model competition and adaptive market responses. The analysis contributes to ongoing work on how AI transforms trading behavior, market quality, and financial stability.