The European Union’s Artificial Intelligence Regulation (AI Regulation), approved by the European Parliament in March 2024 and published in the Official Journal of the European Union on 12 July 2024, represents a watershed moment in AI regulation. While not specifically targeting financial exchanges, this legislation will significantly impact how trading venues approach AI integration in their operations.

Scope and Application

It is important to note that the AI Regulation does not explicitly address algorithmic trading or market surveillance systems. However, financial exchanges need to evaluate how the AI Regulation applies to these areas, especially if classified as high-risk applications under Annex III (Article 2).

AI Applications in Financial Exchanges

Financial exchanges are increasingly exploring AI applications that could fall under the Regulation’s purview. According to EY, “AI systems used in operating and maintaining financial infrastructure considered to be critical also fall under the scope of high-risk AI systems”. For instance, key areas identified by the European Central Bank’s May 2024 Financial Stability Review, such as risk assessment, operational processes and data analytics, are particularly relevant.

Regulatory Requirements for High-Risk AI Systems

If any exchange applications are deemed “high-risk” under the AI Regulation, they would need to comply with stringent requirements, including: 

• Risk Management: Implement robust systems for identifying and mitigating risks (Article 9).
• Data Governance: Ensure high-quality data for AI systems (Article 10).
• Documentation and Record-Keeping: Maintain detailed technical documentation and automatic logs (Articles 11-12).
• Transparency and Human Oversight: Provide clear information on AI system capabilities and limitations, with effective human oversight (Articles 13-14).

Compliance and Oversight

The AI Regulation designates financial service regulators as competent authorities for supervising and implementing the Regulation within their respective competences. For exchanges, this likely means their existing regulators will take on this additional responsibility. 

Implementation Timeline

The AI Regulation will be implemented in stages:

• 1 August 2024: Entry into force
• 2 February 2025: Provisions on banned AI systems apply
• 2 August 2025: Provisions on general purpose AI systems apply
• 2 August 2026: Bulk of obligations, including those for high-risk AI systems, apply
• 2 August 2027: Some obligations for high-risk AI systems in specific sectors apply

Preparing for Compliance

To prepare for the AI Regulation, financial exchanges should:

• Assess current and planned AI systems to identify potential high-risk applications.
• Develop comprehensive AI governance frameworks.
• Invest in staff training to ensure understanding of the Regulation’s requirements.
• Engage proactively with regulators to clarify compliance expectations.

Conclusion

The AI Regulation presents both challenges and opportunities for financial exchanges. It mandates significant resources for compliance but also offers a framework for responsible AI development. Exchanges that proactively adapt to this new regulatory landscape may find themselves better positioned to leverage AI’s potential while maintaining the trust of market participants and regulators.