The European Securities and Markets Authority, or ESMA, is preparing to place artificial intelligence and tokenization at the center of its supervisory focus in 2027. That move signals a clear shift in how European regulators intend to monitor emerging technologies as they become more embedded in financial services. Rather than treating AI and tokenization as niche innovations, ESMA is treating them as key areas of market infrastructure that deserve coordinated oversight across the EU.
What the 2027 supervisory priority means
According to the available information, national regulators will play a central role in the next phase of oversight. The plan involves mapping how firms use AI and tokenization in client-facing activities, conducting checks on a subset of firms, and working toward more common supervisory approaches. In practical terms, this means regulators will not simply issue broad guidance and wait to see what happens. They will look closer at how these technologies are actually being used in real market conditions.
The focus on client-facing uses is especially important. AI is already being used in areas such as customer communication, risk assessment, portfolio management, fraud detection, and automated trading support. Tokenization, meanwhile, is increasingly connected to digital asset issuance, settlement, custody, and new forms of investment products. Because these technologies can affect investors directly, regulators are likely to pay close attention to how firms explain their use of them, how risks are managed, and whether clients receive fair and transparent treatment.
Why AI and tokenization are becoming supervisory priorities
AI is changing how financial firms operate
Artificial intelligence has moved far beyond experimental pilots. Many financial institutions are now using machine learning models to support decision-making, improve efficiency, and reduce operational friction. That creates opportunities, but it also introduces new supervisory questions. For example, how can a firm demonstrate that an AI model used in investment advice or credit assessment is reliable, explainable, and free from harmful bias? How should model risk be monitored over time, especially when market conditions change?
These are not theoretical concerns. If an AI system makes a mistake at scale, the impact can be much broader than a single human error. Regulators are therefore likely to focus on governance, model validation, data quality, accountability, and the ability of firms to explain decisions to clients and supervisory teams.
Tokenization is reshaping asset markets
Tokenization is also becoming a major focus because it can change how assets are issued, traded, and settled. By representing ownership of assets on a digital ledger, tokenization has the potential to improve liquidity, reduce settlement time, and expand access to certain investment opportunities. But it also raises supervisory questions around investor protection, market integrity, cybersecurity, and the legal status of tokenized instruments.
For ESMA, the challenge is to ensure that innovation does not outpace oversight. Tokenized products can cross traditional market boundaries, which makes a coordinated EU-wide approach more relevant than a fragmented national one. If one member state takes a very strict position while another is more permissive, firms and investors could face inconsistent rules, uneven competition, or regulatory arbitrage. That is likely one reason common supervisory approaches are part of the plan.
The role of national regulators
One of the most interesting aspects of the 2027 priority is the emphasis on national regulators working together. ESMA does not replace national supervisors; it coordinates them. In this case, national regulators will be expected to map how local firms are using AI and tokenization, particularly in areas that affect clients. That mapping exercise will help supervisors build a clearer picture of where the technology is being deployed, which sectors are most exposed, and where risks may be concentrated.
The planned checks on a subset of firms also
Related read: Hyperliquid Hits $1.4B in 24-Hour Perp DEX Volume: What the Number Really Tells Us
