Generative and agentic AI are making their way into enterprise production, bringing new questions for Chief Data Officers (CDOs) and other data and AI leaders responsible for making these technologies work at scale.

CDO Magazine’s latest Trend Report, Generative + Agentic AI: Moving From Experimentation to Enterprise Scale, combines original research with perspectives from thought leadership contributors on the opportunities, challenges, and decisions that come with putting GenAI and agentic AI into practice.

Leading the report are research findings from Tuhin Chattopadhyay, Ph.D., based on responses from 118 data and AI leaders. The research examines how far agentic AI has moved into production, where companies are encountering challenges as they move beyond experimentation, and whether governance and control practices are keeping pace with more mature deployments.

While 65.3% of respondents reported reaching early production or beyond, 64.4% had encountered what the research identifies as a “transition wall” between experimentation and production. Among those already in production or beyond, 84.4% reported at least one of the control gaps examined in the research.

From there, our thought leadership contributions begin with the foundation for putting these technologies to work in the enterprise: moving GenAI into production, ensuring the information it relies on remains relevant, and establishing ownership of the context layer.

In “Operationalizing Generative AI: Lessons from Early Enterprise Deployments,” Karthikeyan Ilangovan shares lessons from early enterprise GenAI deployments and considerations for moving projects from pilots into production.

Justin Heller‘s “The Relevance Imperative: Governing the Information AI Relies On” looks at why relevance deserves attention alongside traditional measures of data quality. Murali Kallem examines questions around ownership, accuracy, control, and agility in “Who Owns the Context Layer?: Architecting Accuracy, Control, and Agility in the Era of Agentic AI.”

As AI systems gain greater autonomy, the focus moves to oversight and governance. Tiffany Perkins-Munn, Ph.D., takes a closer look at human-in-the-loop oversight and whether it provides the level of accountability enterprises expect from it in “The Emperor’s New Loop: Why ‘Human-in-the-Loop’ Is the Flimsiest Leaf in Enterprise AI.” Moataz Mahmoud‘s “Agentic AI in the Enterprise: Governing the Shift From Assistance to Execution” examines the boundaries and controls needed as agentic AI moves from providing assistance to taking action.

The final section turns to security. In “Non-Human Identity Security: Managing Access, Control, and Risk in AI-Driven Systems,” Dhivya Nagasubramanian addresses the access and control risks that come with AI-driven systems interacting with enterprise resources.

The report also features curated insights from executives and The Enterprise Stack, a collection of additional resources to continue exploring the topics covered throughout the report.

Many thanks to each of our contributors for the research, expertise, and experiences they brought to it. We are excited to share this Trend Report with you and hope you’ll take some time to explore the findings and hear directly from the people helping shape this work in practice.

Read the full CDO Magazine Trend Report, Generative + Agentic AI: Moving From Experimentation to Enterprise Scale.