Generative AI adoption has become widespread among global knowledge workers, with surveys indicating that between one-third and three-quarters of professionals regularly use tools such as ChatGPT, Copilot, and Gemini in their daily work. In France, 35% of managers report weekly use, with 12% using AI tools almost every day, and younger workers and managers leading adoption rates. In the United States, 43.2% of workers use generative AI at work, applying it to about one-third of their tasks and reporting a tripling of productivity on those tasks. Despite high individual usage and reported productivity gains, organizational support and formal training remain limited. Only 24% of managers in France have received AI training, though 72% express interest, and many users operate without direct company guidance. Additionally, 56% of French managers are open about their use of AI at work. Corporate perspectives on generative AI are evolving, with 76% of large enterprises and 55% of small and medium-sized businesses in France viewing AI as an asset. Globally, 71% of companies reported using generative AI in at least one business area in 2024. Venture capital funding for generative AI surged to $45 billion, nearly doubling from the previous year. However, the transition from experimentation to enterprise-wide value remains slow. According to an IBM study of 2,000 CEOs, only 25% of AI initiatives have delivered expected returns on investment, and just 16% have scaled across the entire organization. About half of CEOs report realizing value from generative AI investments, while many cite disconnected technology and skills gaps as barriers to broader impact. In Brazil, 46% of generative AI projects have successfully moved beyond testing to integration in production processes. Companies there and in other countries are increasingly seeking employees with generative AI expertise. Recent research highlights a shift in how generative AI is used. While technical tasks like coding and content creation remain common, people are increasingly turning to AI for personal support, such as therapy, companionship, and life organization. The rapid evolution of AI models means that selecting the right tool for specific tasks is crucial. Experts like Andrej Karpathy emphasize the importance of matching models—such as GPT-4o for simple tasks and o3 for complex reasoning—to user needs. ChatGPT is approaching 1 billion users, with 2.5% now converted to paid subscriptions.