The AI wave was initially seen as an opportunity for model developers, but recent earnings reports show a cohort of large, established European technology groups are unexpectedly reaping the rewards of artificial intelligence. From SAP (SAP.US), Capgemini, Sopra Steria to OVHcloud, multiple European tech giants are reporting stronger demand, faster growth, or upgraded guidance, all driven by the same factor: businesses are moving from AI experimentation to full-scale deployment.

However, these companies are discovering that generating real productivity from AI within complex organisations is far more difficult than acquiring the technology itself. Large organisations are unlikely to rely on a single AI vendor; instead, they will select different models based on task requirements, performance, security levels, and compliance needs. The current challenge is no longer "which model to choose," but how to seamlessly integrate AI with an organisation's existing software, data, and business processes.

UBS noted in a recent report: "AI application is the real battleground and the core of value creation." This aligns perfectly with the traditional strengths of European legacy tech companies, which have long specialised in helping large organisations integrate complex technologies, well before generative AI emerged. Most large organisations are not starting from a "technical blank slate." AI systems must be compatible with decades of accumulated software stacks, fragmented databases, custom applications, and increasingly stringent governance requirements. They must also access real-time enterprise data, manage permissions, retain audit trails, and embed into workflows employees already use.

The complexity of the task is becoming one of the biggest bottlenecks for AI deployment. Boston Consulting Group points out that the pace of deployment is outstripping corporate management capabilities, with over 70% of investors expressing concern about whether organisations have the technical and operational capabilities needed for AI success. As businesses move from experimentation to practical application, spending on AI implementation, integration, and governance is becoming an increasingly critical part of the value chain.

SAP's cloud backlog grew 26% on a constant currency basis to €22.9 billion, as companies continue to migrate core systems like finance, procurement, supply chain, and human resources to its cloud platform, which is increasingly becoming a foundation for AI deployment. The company's recent acquisitions of data specialist Dremio and AI firm Prior Labs highlight the growing importance of making enterprise data accessible to AI applications.

Capgemini raised its full-year growth target after order bookings grew 9.2%, while Sopra Steria upgraded its outlook after organic growth accelerated to 5.3%. Both Paris-listed companies have benefited from the follow-up work after AI adoption: integrating models into workflows, managing data, and building governance systems. This work is particularly valuable in sectors like defence, aerospace, healthcare, and critical infrastructure, where AI must be embedded into specialised software and highly controlled operational processes.

Another trend is further consolidating the position of established European companies: the increasing demand from businesses for control over AI deployment. Publicis CEO Arthur Sadoun said customers are increasingly wanting to run advanced AI models in environments where they can control their own technology and data. This preference is particularly strong in defence, aerospace, and critical infrastructure, where sovereignty, security, and compliance concerns are paramount.

Airbus’s decision to deploy sensitive industrial and defence applications on the Scaleway cloud platform, owned by French telecoms group Iliad, while using AI tools co-developed with Mistral, is a microcosm of this trend. Airbus expects approximately 70 critical applications to run on Scaleway by the end of 2028. OVHcloud reported a 20.2% increase in public cloud revenue in the third quarter, providing initial evidence that the demand for sovereign, European-controlled AI infrastructure, free from extraterritorial laws like the US Cloud Act, is beginning to translate into commercial growth.

Of course, European legacy tech companies still need to prove that AI-driven demand is sustainable, while ensuring margins can withstand pressure from the automation of low-value consulting and software work. However, the recent earnings results send a clear signal: the biggest beneficiaries of AI may not be limited to model builders. The companies that can make these models "work" within large global organisations are now taking centre stage in this transformation.