Study: AI Operations Continue to Lag High Adoption Rates

While 78% of organizations now use AI in at least one business function, a significant "scaling gap" persists, with nearly two-thirds yet to expand AI across the enterprise. This operational challenge is less about the technology itself and more about people and processes; around 70% of implementation issues are attributed to these areas, with only 10% being related to the AI algorithms. Consequently, confidence in corporate AI strategy has fallen, with a notable drop among CTOs and CEOs. A key hurdle is the lack of AI fluency in leadership, with only 55% of CTOs believing their executive teams fully grasp the associated risks and opportunities. This contributes to a failure to redesign workflows and manage organizational change, which are primary obstacles to achieving return on investment. As a result, 74% of companies report struggling to see tangible value from their AI initiatives. In regulated industries like finance and healthcare, the gap is even more pronounced due to stringent compliance and data privacy requirements. A recent survey shows AI adoption for support operations at 92% in tech companies, compared to just 58% in regulated sectors. These industries require a different, more governance-forward approach to AI, embedding auditability, explainability, and compliance checks throughout the AI lifecycle. To bridge the gap between adoption and operationalization, many are turning to agentic AI architectures and autonomous workflows. These systems are designed to reason, plan, and act autonomously across various tools and platforms, moving beyond simple task automation to handle more complex, multi-step processes. This shift necessitates a move towards goal-oriented, task-centric API design, enabling AI agents to consume meaning and context, not just raw data. The rise of agentic AI elevates the importance of robust governance frameworks, such as the NIST AI Risk Management Framework, to manage the unique risks associated with autonomous systems. For compliance officers, this means developing AI specific compliance programs to monitor AI operations, audit for bias, and safeguard protected data, especially as employees increasingly use generative AI tools. Enterprise architects are now at the forefront of this transition, tasked with redesigning business processes and the underlying technology to support agentic AI. This involves moving away from fine-grained, CRUD-style APIs towards designs that enable autonomous reasoning. The focus is on creating a modular and scalable foundation that allows for the repeatable build-out of AI agents across different business functions. Successful AI operationalization is increasingly seen as a people and data problem, not just a technology one. Enterprises that succeed often dedicate significant resources to their data infrastructure and tackle organizational data silos before scaling AI tools. This highlights a growing understanding that without a solid data foundation and effective change management, even the most advanced AI models will fail to deliver significant business impact.