Artificial intelligence (AI) is no longer a speculative technology in financial markets. It is already embedded across trading, risk management, compliance, client engagement, and operational decision-making (BIS, 2023; McKinsey, 2023). From algorithmic execution and market surveillance to credit scoring and fraud detection, AI promises efficiency gains that were unthinkable a decade ago.
Yet as adoption accelerates, a more difficult question is emerging: does AI merely enhance existing market processes, or is it quietly reshaping market structure itself — and with it, the nature of systemic risk?
The efficiency case is compelling
The productivity benefits of AI are well documented. Machine-learning models can process vast datasets in real time, identify patterns invisible to human analysts, and automate decisions at speed and scale (BIS, 2023). For market participants under pressure to reduce costs, improve execution quality, and strengthen compliance outcomes, AI has become strategically important rather than optional.
In surveillance and risk management, AI techniques improve anomaly detection and reduce false positives relative to traditional rules-based systems (IOSCO, 2021). In trading, AI-driven execution algorithms optimise order routing and liquidity sourcing across fragmented markets, often improving price outcomes for clients.
Used responsibly, these tools increase efficiency and transparency at the firm level.
Where efficiency turns into fragility
Systemic concerns arise not from AI in isolation, but from widespread convergence. As similar models are trained on overlapping datasets and optimised toward similar objectives, behavioural diversity across market participants can diminish (BIS, 2023; IOSCO, 2021).
Financial crises rarely result from isolated failures. They emerge when many actors respond to stress in the same way at the same time. AI risks amplifying this dynamic. During periods of market stress, correlated model behaviour may accelerate liquidity withdrawal, reinforce price dislocations, or overwhelm traditional market-stabilisation mechanisms (BIS, 2023).
Importantly, these vulnerabilities often remain hidden during benign market conditions. Machine-learning models tend to perform best when historical relationships hold and market regimes remain stable. It is precisely during regime shifts — when those relationships break down — that over-reliance on automated decision-making becomes most dangerous.
Governance is lagging adoption
One of the defining challenges of AI adoption in financial markets is that governance frameworks have not kept pace with technological capability. While firms can often describe what their models do, they struggle to explain why models behave in particular ways under stress (ECB, 2024).
Traditional model-risk frameworks were not designed for adaptive systems that evolve over time. Accountability becomes blurred when decisions are automated, distributed, and embedded across complex technology stacks.
As regulators increasingly emphasise explainability, accountability, and senior-management oversight, firms will need to demonstrate not only technical competence but also effective AI governance (Deloitte, 2024; WEF, 2023).
From firm-level risk to market-level responsibility
The next phase of the AI debate will extend beyond internal efficiency to broader market impact. If AI materially influences liquidity provision, price formation, or execution quality, it becomes a market-structure issue, not merely an internal tooling choice (BIS, 2023).
This raises difficult questions:
- Should minimum standards for AI governance apply across market participants?
- How can regulators monitor systemic AI risks without stifling innovation?
- What responsibilities do firms have to ensure optimisation strategies do not collectively destabilise markets?
These questions remain unresolved — but ignoring them would be a mistake.
Regulatory perspective on AI and market risk
Regulators are increasingly alert to the systemic implications of AI adoption in financial markets. At the 2026 FSCA Industry Conference, a dedicated panel on artificial intelligence highlighted concerns around model opacity, over-reliance on automated decision-making, and the potential for correlated behaviour across market participants.
Importantly, the discussion emphasised that governance frameworks are still evolving, and firms will increasingly be expected to demonstrate explainability, accountability, and effective oversight of AI-driven processes, particularly where these influence trading, risk management, or client outcomes. (FSCA, 2026).
A measured path forward
AI is neither an existential threat nor a silver bullet. It is a powerful force multiplier. Used thoughtfully, it enhances market efficiency and resilience. Used blindly, it risks accelerating the very behaviours regulators and practitioners have spent decades trying to control.
The challenge for financial markets is not whether to adopt AI — that decision has already been made. The real test will be whether firms, regulators, and market infrastructures evolve quickly enough to govern it before efficiency gains quietly turn into systemic vulnerabilities.
References
- Bank of International Settlements (BIS). 2023. Artificial intelligence and machine learning in financial services. BIS Annual Economic Report. Available at: https://www.bis.org/publ/arpdf/ar2023e.htm
- Deloitte. 2024. AI governance in financial services. Available at: https://www.deloitte.com/global/en/Industries/financial-services.html
- European Central Bank (ECB). 2024. AI in financial markets. Available at: https://www.ecb.europa.eu/pub/financial-stability/fsr/special/html/index.en.html
- Financial Sector Conduct Authority (FSCA). 2026.
FSCA Industry Conference 2026 - International Organization of Securities Commissions (IOSCO). 2021. Artificial Intelligence and Machine Learning in Capital Markets. Available at: https://www.iosco.org/library/pubdocs/pdf/IOSCOPD684.pdf
- McKinsey & Company. 2023. The state of AI in financial services. Available at: https://www.mckinsey.com/industries/financial-services/our-insights
- World Economic Forum (WEF). 2023. The use of AI in financial markets. Available at: https://www.weforum.org/reports
