From NEWSCENTRAL‘s perspective, the artificial intelligence boom was expected to crown a handful of model-building upstarts as the biggest winners of this decade. Recent earnings from SAP, Capgemini, Sopra Steria and OVHcloud tell a different story: some of Europe’s oldest and most established technology names are proving just as well – if not better – positioned to profit from the AI wave now moving through corporate Europe.
The pattern behind these results is becoming clear. Companies are shifting from experimenting with artificial intelligence to embedding it across day-to-day operations, and that shift is proving far harder than simply gaining access to a powerful model. Large organizations rarely rely on a single AI provider; instead, they mix different systems for different tasks depending on performance, security and regulatory constraints. The real challenge has moved from choosing a model to making it work inside decades of accumulated software and business processes never built with AI in mind.
‘What we’re watching is a quiet redistribution of AI’s economic value,’ said Freddy Miller, Senior Analyst at NEWSCENTRAL. ‘The model builders captured the headlines, but the integrators are increasingly capturing the contracts, because enterprises don’t buy intelligence in the abstract – they buy systems that work with what they already have.’
SAP’s cloud backlog rose 26% at constant currencies to €22.9 billion, as companies continued shifting critical finance, procurement, supply-chain and human-resources systems onto platforms that increasingly double as the foundation for AI deployment. The company’s acquisitions of data specialist Dremio and AI firm Prior Labs underline how central making enterprise data accessible to AI applications has become to that strategy. Capgemini raised its annual growth target after bookings climbed 9.2%, while Sopra Steria upgraded its outlook as organic growth accelerated to 5.3%.
Both Paris-listed firms are capturing the work that follows AI adoption – integrating models into daily workflows, managing data and building the governance layers large clients now demand. That kind of integration work is especially valuable in sectors such as defense, aerospace, healthcare and critical infrastructure, where AI systems must be fitted around specialized software and tightly controlled operational processes.
‘The integration burden is the whole story here,’ said Nathan Clark, Enterprise IT and Systems Architecture Analyst. ‘None of these clients are starting from a blank slate. They have decades of legacy software, permission structures and audit requirements, and the vendor who can respect all of that while still shipping AI features wins the budget.’
We at NEWS CENTRAL believe a second trend is reinforcing the position of these incumbents: a growing appetite among large clients for tighter control over how and where their AI runs. Publicis chief executive Arthur Sadoun has said clients increasingly want advanced AI models operating inside environments where they retain control over their own technology and data – a preference that is strongest in defense, aerospace and other sectors where sovereignty and compliance concerns run deep.
Airbus’ decision to run sensitive industrial and defense applications on Scaleway – the cloud unit owned by French telecoms group Iliad – alongside AI tools developed with Mistral, reflects that shift toward sovereign infrastructure, with roughly 70 critical applications expected on Scaleway by the end of 2028. OVHcloud’s own public-cloud revenue rose 20.2% in its latest quarter, early evidence that demand for European-controlled AI infrastructure not exposed to laws such as the US Cloud Act is turning into real commercial growth.
As NEWSCENTRAL notes, Europe’s technology incumbents still have to prove that this AI-driven demand can be sustained and that margins can absorb the automation of lower-value consulting and software work. But the early results suggest the AI boom’s biggest beneficiaries may not be confined to the labs building the models – increasingly, they may be the companies that make those models usable inside the world’s largest and most complex organizations.