The information technology industry is undergoing a sweeping transformation driven by generative AI, according to GreenSI. This change is not confined to IT departments alone but is affecting a wide range of stakeholders, including software publishers, hosting providers, and service companies. This widespread shift is making it harder to plan and evaluate the return on investment (ROI) for technology projects. Many current projects may become obsolete within six months, and the actual cost of implementing solutions is often difficult to predict and can nearly double. As a result, businesses are increasingly favoring agility, iterative development, and a more cautious approach to long-term projects. For over two decades, four key players in the IT industry—IT departments, SaaS (Software as a Service) publishers, hosting providers, and ESNs (External Service Providers)—operated under a shared economic model that had become unchallenged. However, generative AI is now altering the way value is distributed among these actors. This is not just a technological shift but a fundamental change in the economic structures that have governed the industry for years. As AI reshapes the landscape, the traditional ways of earning revenue and capturing value are being redefined. Major companies are already adapting to this new reality. McKinsey, for example, has deployed nearly 12,000 AI agents to support its project teams, helping with tasks such as creating presentations and verifying data. This has led to a significant reduction in team sizes, from about 14 people to as few as two or seven members. Similarly, Accenture in Australia has restructured its operations by merging five units into a new division called "Reinvention Services," aimed at helping clients navigate the challenges of AI and responding to declining business prospects. Microsoft has also taken steps to regain control over its AI technology, having previously outsourced it to OpenAI. Now, it is developing its own AI models and agents, signaling a shift toward internalization and greater control over its AI-driven services. Generative AI is reshaping four core pillars of the software industry’s economic model. First, the cost of building software is plummeting, prompting IT departments to reconsider whether to develop solutions in-house or outsource them. Second, the cost of using AI services—often measured in "tokens"—is rising rapidly, challenging the traditional SaaS model. Third, the unpredictable nature of AI-driven consumption is making infrastructure costs more volatile, leading to the emergence of a new discipline known as FinOps for AI. Finally, the integration of AI systems is becoming more routine, with autonomous agents handling many tasks, which threatens the traditional role of ESNs that rely on selling installation time. The underlying theme of these changes is clear: the transformation is not just technological, but fundamentally economic. Businesses must recognize that their traditional assumptions about IT management, value creation, and cost structures are being challenged daily. This shift is evident in the U.S., where consulting firms are evolving into platform publishers, and major cloud providers are moving closer to implementation services. The IT department, caught in the middle, is seeing a shift in pricing from time-based models to outcome-based ones. Each actor in the industry is trying to avoid becoming obsolete as the value distribution changes. In France, the impact of these changes is particularly acute. French IT departments are among the most reliant on outsourcing, with over 70% dependence on foreign SaaS, software, and platforms. ESNs, whose business model is based on selling installation hours, are already seeing a decline. In 2025, the domestic IT market saw its first-ever decline, and the upcoming presidential election, with its heavy reliance on public procurement, may not help. The issue of sovereignty—often discussed in geopolitical terms—is now also a practical concern, as questions arise about who controls the development, billing, and usage of AI technologies. GreenSI will continue to explore these changes in future articles, focusing on how IT departments are adapting to the reversal of the "make or buy" equation and the need for new engineering skills. It will also examine the challenges faced by ESNs, the rising cost of SaaS, and the emergence of new disciplines to manage AI-driven infrastructure. Stay tuned for more insights as the landscape continues to evolve.