Europe’s push to reduce its dependence on US technology providers is running up against a practical challenge: sovereign cloud and artificial intelligence services are not yet mature enough for major banks to adopt at scale, according to Temenos CEO Takis Spiliopoulos.
The comments from the Swiss banking software company’s chief executive highlight the gap between Europe’s ambition to gain greater control over its digital infrastructure and the operational requirements of financial institutions that handle sensitive customer information across multiple jurisdictions.
The European Union is pursuing technological sovereignty through proposals including the Cloud and AI Development Act, seeking to reduce reliance on US hyperscalers amid concerns about foreign access to data and the dominance of Amazon, Microsoft and Google in cloud computing.
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
But establishing local data centers does not necessarily resolve the underlying concerns, Spiliopoulos told Reuters. Banks must also understand where their information can ultimately be accessed, under which legal frameworks, and how cloud providers manage data across borders.
“Sovereign data or sovereign AI … it’s not yet at a maturity level that they would really entertain this,” Spiliopoulos said, referring to banks.
He said prospective European sovereign cloud and AI services would need more time to develop into solutions capable of meeting the requirements of large financial institutions.
For an industry in which data protection, operational resilience and regulatory compliance are central to technology procurement, the distinction is important. A cloud service may store information within Europe but still leave questions about who can access it, which laws apply to the provider, and how information is handled across international operations.
European governments and regulators have increasingly focused on technological dependence as cloud computing and advanced AI become more important to businesses and public services. The concentration of infrastructure among a small number of US technology companies has raised concerns about Europe’s ability to control critical digital systems and protect sensitive information from access under foreign legal authorities.
Sovereign cloud and AI initiatives seek to address these issues by giving European organizations greater control over data, infrastructure, governance and technology deployment. However, sovereignty can mean different things in practice, ranging from the physical location of data to the ownership of infrastructure, control of encryption keys and the legal jurisdiction governing a service provider.
Spiliopoulos indicates that banks are looking beyond the location of servers when assessing whether a service qualifies as sufficiently sovereign for their needs.
For financial institutions, the challenge is deemed complex because they must manage confidential customer records, payment information and other sensitive data while meeting regulatory obligations in the countries where they operate. A bank serving customers across Europe may need a technology arrangement that satisfies different national requirements without fragmenting its systems or making cross-border operations more difficult.
The situation has created a demanding set of requirements for emerging European cloud providers. They must offer not only assurances about data control but also the reliability, security, performance, and integration capabilities needed by large banks.
The scale of the task is considerable. Major financial institutions rely on technology systems that support transactions, risk management, customer services, and regulatory reporting. Replacing or migrating these systems requires extensive testing and careful management of operational risks. A provider that meets sovereignty requirements but cannot support those workloads consistently may struggle to win large banking contracts.
The availability of local data centers is therefore only one part of the solution. Banks also need clarity about contractual protections, access controls, incident response, service continuity, and the treatment of data when services depend on infrastructure or personnel outside a particular jurisdiction.
Cross-Border Banking Complicates Europe’s Technology Ambitions
Spiliopoulos also identified cross-border operations and regulatory requirements as obstacles to the adoption of sovereign cloud and AI services, particularly among banks operating across multiple jurisdictions.
This exposes a tension at the heart of Europe’s digital policy. The EU wants to strengthen its technological autonomy, but large banks also need technology platforms that can operate consistently across national borders. If sovereign services are designed around fragmented national requirements, institutions could face additional complexity in managing their systems.
Banks will need to assess whether emerging offerings can accommodate their existing infrastructure, comply with applicable rules and support the integration of services across markets. They must also consider the commercial implications of moving away from established cloud providers, including migration costs, performance differences and the availability of specialized AI tools.
The challenge is not necessarily that European providers cannot develop suitable services, but that building the necessary capabilities takes time. Large banks generally cannot migrate critical systems on the strength of policy objectives alone. They need evidence that a provider can meet operational and regulatory requirements over the long term.
The same considerations apply to AI. Banks exploring AI applications must evaluate not only the quality of models but also the handling of sensitive information, the ability to control how systems are deployed, and the reliability of the infrastructure supporting them. Sovereign AI offerings will need to address these requirements alongside the technical demands of running advanced models.
The situation has birthed a commercial test for Europe, about its sovereignty initiatives producing services that institutions can use without sacrificing the scale, performance and integration they expect from existing providers.
The region’s dependence on US hyperscalers is unlikely to be resolved simply by encouraging local infrastructure investment. European alternatives must demonstrate that they can support demanding workloads and provide credible guarantees about data governance and legal control.
Temenos’s assessment points to a distinction between political ambition and procurement decisions. Governments can establish policy frameworks and encourage domestic investment, but banks will adopt new platforms when they are satisfied that the services meet their technical, legal, and operational requirements.