In 2025, agentic artificial intelligence continued to expand in companies, heralding a new era of automation and process transformation. Although implementation in organizations remains largely experimental, this year has seen increased interest in agentic AI, which is capable of planning and executing tasks autonomously with “agents” that require little human intervention.
It has been three years since generative AI arrived in organizations, and its progress has been uneven. Half of the entities intend to use AI to transform their businesses and redesign workflows, and among those already implementing innovations, most are still in the experimentation or pilot phase with a view to scaling up.
The McKinsey & Company survey, conducted among nearly 2,000 companies of all sizes and across all sectors in 105 countries, found several relevant facts about the global state of AI in 2025:
– The challenge of scaling AI and implementing agents: 88% of respondents said they use AI in at least one business function (up from 78% in 2024), and only a third have begun to scale these processes. As for experimentation with agents, capable of acting by planning and executing multiple processes in a workflow. Twenty-three percent of respondents are already scaling to an agentic AI system in one or two functions, especially in IT and knowledge management areas.
– The impact on jobs and the search for hybrid intelligence. Respondents shared different perspectives on how AI could affect the size of the workforce in 2026. In general, little or no change in the number of employees is expected. Leaders rarely see AI as a standalone solution, and their challenge is to create a “hybrid intelligence” that generates interaction between experienced people and AI solutions. A few even reported that AI led them to increase their workforce. Software and data engineers are in the highest demand.
Among those expecting changes in employee numbers, 17% of respondents report decreases in headcount over the past year and 30% in 2026 as a result of AI use. This trend is more common in large organizations.
– Dealing with the risks of AI is becoming increasingly common. There has been an increase in the number of organizations surveyed that are mitigating the risks associated with AI, such as privacy, organizational reputation, and regulatory compliance. Inaccuracy is one of the risks that most respondents say their organizations are trying to mitigate.
Most companies say that the use of AI is improving innovation, with almost half reporting improvements in customer satisfaction and competitive advantages. However, the biggest challenge in implementing AI is scalability, especially for smaller companies, compounded by the possibility of closing the gap between implementation and results and significantly impacting corporate profits.
