The ROI of AI 2025

The document "The ROI of AI 2025," developed by Google Cloud based on a survey of over 3,400 business leaders worldwide, examines how organizations are achieving real returns with artificial intelligence, particularly through incorporating AI agents into their processes. The text starts from a central idea: after the expectation phase around generative artificial intelligence, attention now focuses on how to obtain value and tangible results through its practical application.

The report describes the transition from predictive and generative models toward the agentic era, where AI agents can execute tasks and make decisions under human supervision. These agents, conceived as systems that combine advanced models with access to tools and data, are already deployed across numerous sectors and organizations. Companies that adopted this technology earlier report clear benefits: increased team productivity, better customer service, process optimization in areas like marketing, human resources and IT, and support for product and service innovation.

One of the main findings is that early adopters of AI agents dedicate a significant portion of their budgets to these solutions and, consequently, show more consistent and faster returns. The document identifies five key areas where artificial intelligence is generating greater returns: employee and process productivity, customer experience improvement, business growth, marketing, and security strengthening. These areas serve as reference for other companies seeking to guide their investments and prioritize impactful use cases.

The analysis also includes comparisons between industries and organization sizes, showing that AI agent adoption is spreading widely across all types of sectors, from retail and manufacturing to financial services, healthcare, telecommunications, and the public sector. Although figures vary, the conclusion is that AI agent use is not limited to a specific field, but is shaping a cross-cutting change in how companies work.

The study also highlights the decisive role of senior management. Executive team involvement is a key factor for artificial intelligence projects to succeed, secure funding, and integrate into business strategy. In this regard, the report proposes an action plan that includes the need for executive sponsorship, establishing solid data governance and security frameworks from the start, demonstrating AI value to justify dedicated budgets, granting agents secure and controlled access to internal systems, creating clear usage and compliance policies for orderly scaling, starting with projects that offer visible and measurable benefits, and investing in training and talent development to consolidate internal capabilities.

Finally, the document warns that AI agent adoption still faces challenges. System integration, regulatory compliance, and data protection are critical aspects requiring attention from the beginning. Overcoming these challenges is essential for scaling the technology and consolidating its benefits.

Overall, the report offers a complete and accessible overview of artificial intelligence's state in 2025, aimed at business leaders and technology strategy managers who want to understand how AI agents are transforming work, what the most visible benefits are, and what conditions favor solid and sustainable return on investment.

Key Points

• AI agents mark a new stage in artificial intelligence evolution, moving from generative to agentic.
• Companies adopting this technology early achieve faster and more consistent returns.
• Five areas concentrate the greatest impact: productivity, customer experience, growth, marketing, and security.
• AI agent adoption is cross-cutting and observed across multiple industries and organization sizes.
• Senior management support is decisive for ensuring funding, strategy, and project success.
• The report recommends prioritizing data governance, security, high-impact projects, and staff training.
• System integration, data protection, and regulatory compliance are the main current challenges.
• The document offers an action plan to convert AI into tangible and sustainable results over time.