Enterprise AI outcomes depend on the art of unlearning

Artificial intelligence offers great promise for businesses, yet its successful deployment within large organizations often falls short of theoretical potential. A Gallup report from 2026 highlights a significant disconnect between employee perception of AI's impact and the actual transformation it brings to work. The author, with experience in tech transformations from both an executive and consulting perspective, emphasizes that the key to enterprise AI success lies in empowering employees to unlearn outdated tools and habits, enabling them to embrace new capabilities with confidence.

As the industry moves from passive AI assistants to autonomous AI agents, the human-AI connection becomes a crucial factor. Traditional software delivery models, however, overlook this connection, leading to challenges in scaling AI outcomes. To illustrate this, the author recounts a data migration project for a global logistics company that originally focused on technical complexities but found the most significant hurdle was helping long-tenured business analysts unlearn established workflows.

By pairing technical migration with targeted change management initiatives, the team successfully guided analysts to understand the value of change, accelerating the deployment timeline more than the technical migration alone. The author identifies a fundamental flaw in traditional AI adoption: when employees shift from controlling task-based tools to supervising autonomous agents, a vulnerability gap emerges.

Traditional IT structures, often siloed, struggle to adapt to this environment, leaving frontline staff between old habits and new autonomous capabilities. The solution lies in adopting a modern delivery model centered around human-centered collaboration, with two key roles: the Forward Deployed Engineer (FDE) and the Agentic Design Lead.

FDEs work directly within business processes, collaborating with AI agents to swiftly transform operational goals into production-ready systems. Meanwhile, Agentic Design Leads focus on the human-AI connection and organizational alignment, ensuring that user behavior informs the design process and promoting a seamless human-AI connection throughout the solution.

To safely scale agentic workflows, organizations must establish clear risk categories: deterministic environments, where errors may pose physical or human safety risks, and probabilistic environments, where autonomous agents can excel. By dividing operations between these categories, leaders can implement appropriate safeguards and enable cross-functional teams to experiment in secure sandboxes, ultimately guiding employees through a grassroots marketing campaign of internal change management initiatives.