Orange Business offers end-to-end digital transformation specifically to global enterprises. What are your key verticals and primary regions in terms of demand for your services?
Orange Business is the B2B division of Orange, which has a revenue of approximately €40 billion. Orange Business itself generates close to €8 billion annually. We serve all sizes of enterprises, from small and medium-sized (SMEs) in France , as well as the public sector. Outside of France, we focus primarily on multinational companies and operate in 65 countries, serving customers across more than 200 territories. Unlike other telecom companies, we have diversified into cybersecurity, cloud, and digital services, which are now billion-euro businesses for us. Our strongest verticals include industrial manufacturing, financial services, and logistics. Recently, we have also made significant investments in defense and healthcare, particularly in France, driven by geopolitical factors and the growing need for digital sovereignty.
How do you perceive the state of competition in the digital services industry today, particularly with the growing emphasis on digital transformation and AI?
Digital transformation has been a focus for over a decade, going through phases like digitalization, cloud adoption, the impact of COVID-19, and now AI, which is the biggest disruption. While many companies are emphasizing AI as a major factor for transformation, it is also disrupting established players, including SaaS vendors in Silicon Valley, who once led the cloud revolution. In today’s market, trust and sovereignty are crucial. AI, being harder to control, has significant geopolitical implications, especially around data production, storage, and management. The challenge now is not just implementing AI, but ensuring it delivers tangible business outcomes. Companies are increasingly looking for partners who can provide secure technology and deliver real, measurable results.
What are some of the biggest misconceptions at the executive level of global enterprises about the return on investment for AI deployments?
One misconception is that AI is a quick fix that will deliver immediate results. Many clients face top-down pressure from their boards to adopt AI, but the outcomes often do not meet expectations. At a recent customer advisory board meeting, CxOs from major multinationals expressed frustration with AI solutions from vendors that promised quick returns but did not deliver.
From our perspective, AI value creation follows a pyramid structure. The first level involves automating simple tasks, which does not result in immediate financial ROI. The second level focuses on process transformation, improving productivity—like making software development 30% more efficient. At the top of the pyramid, AI can handle core processes, transform business models and open new markets, but this requires integration, process change, and skill development, which is where managed services and integration support come into play.
AI is not an instant solution; it needs careful, structured transformation to generate true value.
You described industrial manufacturing, financial services, logistics, defense, and healthcare as some of your key verticals. Which of those sectors are achieving top-of-the-pyramid ROI already?
In the life sciences industry, for example, AI is already significantly impacting drug discovery and development. Traditionally, pharmaceutical companies had long, investment-heavy cycles, but AI is accelerating this process. If a startup uses AI to speed up drug testing, it is a major game-changer. In regulated industries like defense and healthcare, core processes can be significantly improved with AI, particularly in compliance and verification management. AI can handle large volumes of data, streamlining these complex processes. However, while AI can enhance these tasks, trust is essential. There needs to be confidence that the AI system will work securely and reliably.
Trust is critical in regulated industries like healthcare and defense, where data sovereignty is paramount. In France, we provide SecNum Cloud, a private cloud platform that complies with local sovereignty requirements and adheres to Europe’s strictest security standard. Clients can use this platform to run AI models securely, ensuring sensitive data is handled in line with local regulations. For non-sensitive data, like general email summaries, clients may use public cloud services. But for sensitive data, they rely on the trusted cloud. This flexible approach allows companies to balance security and sovereignty, ensuring that different levels of data sensitivity are handled appropriately.
Speaking of private cloud platforms, do you think the role of hyperscalers could be challenged by a more pluralistic hosting ecosystem, or will they continue to dominate?
Yes, hyperscalers will remain dominant, especially in areas like large language models (LLMs) and compute power, but their role is evolving. The geopolitical landscape is changing, and it is no longer enough for hyperscalers to have a central cloud presence serving global markets. Companies now require localized compliance, whether in Europe, the U.S., China, or the Middle East. Hyperscalers are responding by strengthening their local presence, especially in regions like Africa, where local infrastructure is crucial for both accessibility and regulatory compliance. This trend suggests a move toward more partnerships and localized deployments, creating a more pluralistic ecosystem. Companies are now seeking partners who can navigate diverse regulatory environments, which is prompting hyperscalers to adapt.
Orange has a collaboration with OpenAI and Meta to support the development of AI models based on regional African languages. What are some of the benefits these language models could bring to IT, businesses, startups, or the broader public service ecosystem?
Africa is underrepresented online, with limited access to the internet and local language content. Current AI models often perform poorly in African languages, which limits their effectiveness. By collaborating with Meta to build language models that reflect African realities, we aim to improve AI accessibility and performance in these languages. This initiative also helps build a corpus of information that will support better AI solutions in the future. For Orange, this investment in AI models for Africa aligns with our values and broader goals. While the immediate ROI might not be evident, this will contribute to the continent's digital transformation, benefiting both societies and businesses across Africa.
What do you consider the most pressing debate in AI-enabled digital transformation services?
The two most pressing debates are ROI at scale and trust. Many companies claim to have achieved ROI with AI, but they often do so with an AI-first approach, without legacy systems to transform. For large companies with established systems, scaling ROI sustainably is still a challenge. On the other hand, trust is crucial as AI becomes more integral to business operations. As AI systems grow more complex, it is essential that companies trust them to operate securely, ethically, and in compliance with regulations. The challenge is leveraging the full potential of AI while ensuring it aligns with company values and operates predictably.
