Airbus Migrates Critical Applications from AWS to Scaleway for Digital Sovereignty

Airbus Migrates Critical Applications from AWS to Scaleway for Digital Sovereignty

Airbus is repatriating its critical digital infrastructure to Europe to mitigate risks associated with non-EU data jurisdiction and ensure digital sovereignty. The aerospace giant is migrating approximately 900 applications from Amazon Web Services (AWS) to Scaleway, a French cloud services provider, with an initial priority group of 70 critical applications.

Strategic Shift to European Cloud Infrastructure

Airbus is prioritizing the migration of its most sensitive systems to put critical data under European control. The migration target includes essential business functions such as:

  • Enterprise Resource Planning (ERP)
  • Customer Relationship Management (CRM)
  • Manufacturing Systems

This move signals that digital sovereignty has evolved from a regulatory preference to a primary commercial driver. By moving these "crown jewels" to a European provider, Airbus aims to protect its intellectual property and operational data from the legal reach and surveillance capabilities of non-European states.

Remaining Dependencies on US Software Ecosystems

Despite the migration of infrastructure (IaaS/PaaS), Airbus remains heavily dependent on American software for productivity and supply chain coordination. This creates a fragmented sovereignty posture:

Supply Chain Management

Airbus manages a global network of 18,000 suppliers. Because the vast majority of these suppliers utilize Microsoft ecosystems, moving supply chain management to a sovereign EU portal does not eliminate the risk, as the data continues to reside on the suppliers' US-based systems.

Productivity and Collaboration Tools

Airbus continues to use Microsoft and Google for productivity applications. The lack of a "blue-chip" European alternative capable of handling large-scale enterprise requirements—specifically those that can respond to competitive tenders with a proven track record of managing massive projects—makes these tools nearly impossible to replace in the short term.

AI Strategy and On-Premises Deployment

Airbus is intentionally avoiding the integration of frontier AI models from US providers into its critical systems. The company's six-pronged AI strategy focuses on areas that can be built and deployed on-premises or within sovereign clouds, including:

  • Knowledge extraction
  • Support chatbots
  • Decision support
  • Flight automation

By avoiding reliance on external agents or code generation from providers like Anthropic (Claude) for critical aircraft systems, Airbus maintains strict control over the logic and security of its flight-critical software.

Industry Implications and Market Trends

The Airbus migration reflects a broader shift in how large enterprises view cloud infrastructure and geopolitical risk.

The Commoditization of Cloud Infrastructure

Industry observers note that cloud infrastructure is becoming increasingly commoditized. This "substrate-independence" allows companies to move workloads based on cost, support, and regulatory compliance rather than technical lock-in. Some organizations are also revisiting hybrid or on-premises models as compute density and liquid cooling reduce the physical footprint required for private data centers.

Impact on US Small Businesses

There is emerging evidence that the push for digital sovereignty is impacting US-based SMEs. Some US entrepreneurs report losing European customers to EU-based competitors, not due to product inferiority, but because European management now mandates the use of EU suppliers to align with foreign policy and data sovereignty goals.

The Challenge of Vendor Lock-in

While IaaS migration is possible, the "sticky" nature of SaaS (Software as a Service) remains a significant hurdle. The industry consensus suggests that designing portable architectures from the outset is becoming a competitive advantage to avoid the long-term risks of vendor lock-in, especially as AI workloads increase the complexity of data dependencies.

Sources