AI in Business: Integration and Governance Under Scrutiny

The Growing Impact of AI in Businesses

Artificial intelligence (AI) continues to transform the business landscape, with an increasing number of companies moving from pilot projects to large-scale production. Companies are no longer just discussing the possibilities offered by AI; they are actively reallocating their budgets and resources to make this transition a reality. Agentic AI, in particular, is generating growing interest due to its potential to automate more processes. However, despite this enthusiasm, many companies still face challenges in achieving full operational success.

Challenges of Large-Scale Adoption

While AI experimentation is widespread, enterprise-wide adoption remains a challenge. Without effective integration of data and systems, reliable automated workflows, and strong governance models, AI initiatives risk getting stuck in the pilot project phase. The rise of agentic AI and the increasing autonomy of models make a holistic approach to data, applications, and systems integration more crucial than ever. Without this, enterprise AI projects are likely to fail.

Predictions and Concerns

Gartner predicts that by 2027, more than 40% of agentic AI projects will be canceled due to high costs, inaccuracies, and governance challenges. The real issue does not lie within AI itself, but rather in the absence of a solid operational foundation that would support these initiatives.

Survey on AI Practices

To better understand how organizations structure their AI operations and deploy successful projects, MIT Technology Review Insights conducted a survey of 500 senior IT executives in mid-sized to large companies in the United States. All of these companies are pursuing AI in one way or another. The results of this survey, complemented by a series of interviews with experts, were collected in December 2025.

Importance of Strong Integration

The findings show that a solid integration foundation is associated with more advanced AI implementations, fostering enterprise-wide initiatives. As AI technologies and applications evolve and multiply, an integration platform can help organizations avoid duplication of efforts and silos while maintaining clear oversight as they navigate the increasing autonomy of workflows.

Key Findings from the Study

Adoption and Success of AI

In recent years, many studies have revealed a lack of tangible success in AI. However, current research indicates that 76% of surveyed companies have at least one department with a fully operational AI workflow. This represents a significant advancement compared to previous years.

Well-Defined Processes and Dedicated Teams

Nearly 43% of organizations succeed with AI implementations applied to well-defined and automated processes. A quarter of companies succeed with new processes, and a third apply AI to various processes. However, two-thirds of organizations lack dedicated AI teams. Only one-third (34%) of organizations have a team specifically responsible for maintaining AI workflows. One-fifth (21%) indicate that central IT is responsible for the ongoing maintenance of AI, while 25% state that the responsibility falls to departmental operations. For 19% of organizations, the responsibility is shared.

Role of Integration Platforms

Companies with enterprise-wide integration platforms are five times more likely to use a wider variety of data sources in their AI workflows. Indeed, 59% of them use five or more data sources, compared to only 11% of organizations using integration for specific workflows, or 0% of those without an integration platform. Organizations using integration platforms also have a more multi-departmental implementation of AI, greater autonomy in AI workflows, and more confidence in granting autonomy in the future.