The European Securities and Markets Authority announced on 23 September that it will launch a Union Strategic Supervisory Priority on digital innovation in 2027. Its initial focus is how supervised entities use AI and tokenisation, with national competent authorities. ESMA: new supervisory priority on digital innovation from 2027
Follow the evidence
Trace how the event could reach markets, then inspect a competing explanation.
Compare explanations
Switch lenses to see what each account explains—and what remains uncertain.
Firms may face more coordinated questions about how AI and tokenisation are governed in supervised activities.
Firms may face more coordinated questions about how AI and tokenisation are governed in supervised activities.
The initial objective may be building regulator expertise rather than imposing immediate new firm-level rules.
ESMA will coordinate supervisory attention across the Union
ESMA said the new Union Strategic Supervisory Priority will begin in 2027 and aims to help supervisors build expertise and capacity for overseeing new technologies. The initial areas are artificial intelligence and tokenisation at supervised entities. ESMA said it would work with national competent authorities and remain flexible as other technologies develop. ESMA: new supervisory priority on digital innovation from 2027
The authority described this as a convergence tool: a way to direct supervisory resources toward areas it considers strategically important across the EU. It will run alongside the existing priority on cyber and operational resilience. The announcement establishes supervisory focus, not a standalone product rule, technology ban or approval for specific AI systems. ESMA: new supervisory priority on digital innovation from 2027
The press release does not specify a common inspection timetable or a single compliance checklist for every firm. Its wording leaves room for supervisors to address use cases as they emerge. For brokerages, trading venues and investment-service providers, the immediate takeaway is to understand where AI and tokenisation already sit in the operating model.
Supervision may look through the technology label to the control chain
An AI feature can influence client communications, suitability processes, order handling, fraud detection or internal risk controls. A governance review should map the system’s purpose, data inputs, human oversight, vendor dependencies and failure response. ESMA’s announcement does not list these as new binding duties, but those elements help firms explain what the technology does and how it is controlled.
Tokenisation may affect ownership records, custody, settlement, reconciliation and the legal treatment of an asset. For a broker or venue, the important question is how the token maps to the underlying right and which entity is responsible at each stage. Supervisors’ interest in tokenisation makes it prudent to keep technical design linked to existing investor-protection obligations. ESMA: new supervisory priority on digital innovation from 2027
Cross-border firms may encounter different use cases and supervisory practices even within a common EU framework. ESMA’s stated goal of coordination could encourage more consistent questions over time, but the release does not guarantee identical interpretations. Firms should maintain a central inventory while preserving jurisdiction-specific permissions and local supervisory communications.
Build an evidence trail before a supervisor asks for one
A practical readiness step is to list AI and tokenisation systems used by the firm or its critical providers, including pilots. Record purpose, customer impact, data access, accountable owners, fallback procedures and independent testing. This is sound operational documentation; it should not be misrepresented as an ESMA checklist or a newly effective regulation. ESMA: new supervisory priority on digital innovation from 2027
Look for supporting factsheets, supervisory convergence work and national-authority guidance after the priority begins. A Union-level initiative can shape supervisory questions before legislation changes, so firms should separate three things in their tracking: current binding requirements, published guidance and future supervisory work programmes.
An alternative reading is that the priority chiefly builds supervisory expertise, with no immediate change to a firm that already has robust controls. That is possible, but the authority has singled out AI and tokenisation as areas for closer coordination. The best response is proportionate preparation, then adjustment when specific supervisory expectations are published.