AFM Urges Strong Human Oversight as AI Drives Faster, More Complex Market Behavior

Key Takeaways

• Unintended Market Coordination: AI-driven trading systems can mirror each other’s behavior, creating price movements that resemble coordination without any actual communication or intent.
• Data Becomes the Weak Link: As models increasingly rely on news, sentiment, and social media, manipulation risks shift from trades themselves to the information feeding algorithms.
• Multi-Agent Systems Add Complexity: The rise of interacting AI agents introduces new layers of unpredictability, making outcomes harder to trace and control.
• Accountability Remains a Human Responsibility: Regulators are emphasizing that firms must maintain clear oversight, governance, and responsibility, even as systems become more autonomous.
• Supervision and Regulation Are Catching Up: Authorities like the AFM are strengthening oversight and working with global peers to address emerging risks tied to AI in trading.

Deep Dive

The rapid integration of artificial intelligence into capital markets is actively reshaping how trading operates today. But as speed and efficiency increase, so too do the risks. In a new report, AI in Capital Markets: Balancing Innovation and Integrity, the Autoriteit Financiële Markten (AFM) makes a clear case that human oversight and accountability must remain firmly embedded in the system.

The regulator’s message is not anti-innovation. If anything, it acknowledges that AI is already delivering tangible benefits, such as faster analysis, lower costs, and more efficient trade execution. The concern lies in how these systems behave at scale, particularly when autonomy begins to outpace human understanding.

One of the more striking warnings in the report centers on how AI-driven trading systems can influence price formation without any explicit coordination. Self-learning models, often trained on similar datasets and pursuing comparable objectives, can begin to mirror each other’s behavior.

The result is a kind of unintended synchronization. Patterns may emerge that resemble collusion, even when no communication or intent exists between market participants. In these conditions, a single model’s misstep, whether due to flawed data or a defensive reaction, can cascade rapidly across the market, distorting price movements in ways that no longer reflect underlying fundamentals.

For regulators and firms alike, this raises a difficult question. If no one intended the outcome, who is responsible for it?

A New Attack Surface in Market Information

The AFM also highlights a shift that will be familiar to anyone watching the evolution of AI systems, data is becoming the primary vulnerability.

Modern trading algorithms increasingly ingest not just market prices, but also news, sentiment indicators, and social media signals. That creates a new risk dynamic. Instead of manipulating trades directly, bad actors could target the information layer that feeds these systems.

A single fabricated or misleading report, if widely disseminated, could trigger a chain reaction across hundreds of algorithms within seconds. The manipulation would not occur through traditional market mechanisms, but through the inputs that guide automated decision-making.

For compliance and risk teams, this reframes market abuse. Surveillance may need to extend beyond trading activity and into the integrity of data sources themselves.

The Rise of Multi-Agent Systems

Alongside individual AI models, the AFM points to the growing presence of multi-agent systems, networks of interacting algorithms capable of coordinating actions in real time.

These systems promise greater efficiency and adaptability, but they also introduce a new layer of complexity. Interactions between agents can produce outcomes that are difficult to predict or control, particularly in fast-moving markets.

The regulator stops short of prescribing specific solutions, but its expectations are clear. Firms deploying these technologies must define boundaries, implement robust controls, and ensure that human oversight remains visible and effective. Responsibility cannot be abstracted away to the algorithm.

Keeping Humans in the Loop

Despite the risks, the AFM notes that the industry is not approaching AI blindly. Market participants are increasingly aware of both the opportunities and the potential pitfalls.

What matters now is execution. Firms must remain agile enough to integrate new technologies while maintaining safeguards that keep risks contained. That includes clear governance structures, defined accountability, and the ability to intervene when systems behave unexpectedly.

The regulator, for its part, is stepping up its own capabilities. It is investing in supervisory expertise, refining expectations around algorithmic trading and AI, and working with international counterparts to ensure a coordinated approach.

The broader goal is not to slow innovation, but to ensure that it unfolds within a framework that preserves trust.

The GRC Report is your premier destination for the latest in governance, risk, and compliance news. As your reliable source for comprehensive coverage, we ensure you stay informed and ready to navigate the dynamic landscape of GRC. Beyond being a news source, the GRC Report represents a thriving community of professionals who, like you, are dedicated to GRC excellence. Explore our insightful articles and breaking news, and actively participate in the conversation to enhance your GRC journey.