2026-04-13

Human Involvement and Clear Responsibility Crucial for AI Deployment in Capital Markets

The Dutch Authority for the Financial Markets (AFM) warns that rapidly advancing AI and multi-agent systems in capital markets can autonomously steer price formation and amplify vulnerabilities through shared data and social media manipulation. The regulator requires market participants to maintain transparent decision-making, visible human oversight, and clear accountability to prevent uncoordinated algorithmic collusion and information-based market distortions. To address these emerging risks, the AFM is strengthening its supervisory framework, tightening rules on algorithmic trading, and collaborating with national and international partners to ensure innovation remains balanced with market integrity.

Autoriteit Financiele Markten logo

Netherlands

Autoriteit Financiele Markten

Click to view thumbnail

News

13/04/26

AI is rapidly changing how trading on capital markets works. This offers opportunities: analyses become faster, costs decrease, and orders are executed more intelligently. However, the same technology exacerbates existing vulnerabilities and can also create new risks. In the report AI in Capital Markets: Balancing Innovation and Integrity, the AFM outlines a future where AI models can operate so quickly and autonomously that they influence market behavior in ways investors do not always fully understand. Therefore, we are now initiating a dialogue with the market and with national and international regulators regarding these opportunities and risks.

In Brief

AI can steer price formation without collusion

Information as a new vulnerability

Rise of multi-agent technology

Looking Ahead

AI can steer price formation without collusion Through self-learning trading systems, patterns may emerge that resemble coordinated behavior, even without parties having consulted each other or intending to do so. These systems can mirror each other's price movements and amplify market trends. Because many models rely on the same data and aim for the same objectives, the market becomes more vulnerable. A faulty or defensive reaction from a single model can spread rapidly in the absence of adequate controls, creating price patterns that are no longer the result of healthy market functioning.

Information as a new vulnerability An increasing number of trading models process not only prices, but also news, market sentiment, and information from social media. Consequently, manipulation can shift from the market itself to the information these models rely on. A single cleverly disseminated fake message could theoretically set hundreds of algorithms in motion within seconds, without a single misleading order needing to be placed.

Rise of multi-agent technology At the same time, alongside independent AI systems, we are seeing the emergence of multi-agent technology. This development offers opportunities for greater speed and efficiency, but primarily demands careful design and appropriate risk management. Market participants must be agile enough to respond to new applications. They determine themselves where technology is deployed, what boundaries apply, and how they monitor risks and potential interventions. Through transparent choices, visible human involvement, and clear responsibilities, trust in the market remains secured.

Looking Ahead The sector is demonstrating a conscious and active approach to the increasing use of AI. In a rapidly evolving technological environment, it is crucial that participants remain agile so that new developments can be integrated in a timely manner. At the same time, appropriate safeguards are essential to keep risks under control and maintain oversight.

We are supporting this development by further strengthening our supervision and internal expertise. Risks are translated into appropriate supervisory measures at an early stage, with attention to the potential need for additional interpretations or legislation and regulations. Recently, the rules and expectations for algorithmic trading and the use of AI have also been tightened. Through continuous dialogue with the sector and in collaboration with national and international partners, we are contributing to a stable and future-proof market where innovation and control are balanced.

Tags

Banks

Investment Institutions

Investment Firms

Issuers

Trading and Settlement Platforms

More Information

'AI in Capital Markets: Balancing Innovation and Integrity' (in English) (pdf, 700 kB)

Public version 'AI in Capital Markets: Balancing Innovation and Integrity' (pdf, 2.3 MB)

Contact regarding this article

Would you like to receive the latest news from the AFM?

Then sign up for our newsletter, and we will keep you informed.

Read more