Broken Processes, Shattered Dreams: Overcoming AI's Hidden Hurdle

The AI hypecycle is monopolized by the major foundational LLMs shipping an endless daily stream of jaw-dropping demos, blockbuster acquihires, and the perfect promise of an automated future. Meanwhile, VC-backed application layer wrappers are marketing for the attention of executives chasing quick wins for fear of being left behind as competitors tout headline-worthy productivity gains.

But inside the walls of some of the world's largest enterprises, a different reality is unfolding. The race to deploy AI isn't a sprint to the shiniest new tool; it's a foundational slog through decades of technical debt, disconnected data systems, and broken business processes. Before the magic can happen, someone has to fix the plumbing.

This is the world of Sergio Gutiérrez-Montero, Director of AI Strategy & Implementation at the biopharmaceutical giant AbbVie. With a focus on the engine room of the enterprise—HR and corporate functions—he tackles the complex, human-centric challenges of using AI for everything from crafting employee succession plans to customizing marketing. He offers a pragmatic reality check for leaders who think transformation is just a software update away.

• A dose of reality: "People think AI is something that you just sprinkle on top," Gutiérrez-Montero says, cutting to the core of the most common executive misconception. The truth, he argues, is that the real work begins with the unglamorous but non-negotiable foundation of data infrastructure.
• Data in all the right places: Before any advanced AI can deliver on its promise, companies must first embark on the painstaking work of connecting their disparate systems of record—the Workdays and ServiceNows where corporate knowledge lives. "The first step is actually readying your data for AI and your infrastructure to make sure that data is in the right places."

But even with perfect data, Gutiérrez-Montero warns that leaders are confronting a deeper problem: the business process itself. Enterprise AI is fundamentally different from the consumer apps we use every day, and most corporate adoption is still in its infancy. "For now," he cautions, "we are still in the phase of 'let's just chat with this drafting tool.' It's still very nascent."

• Broken by design: To move beyond simple drafting tools, companies must first map the complex human workflows that underpin their operations. "When it comes to enterprise AI, you're actually needing to map the underlying workflow—who gets to do what in the process," he stresses. "The process may be broken in the first place." Simply automating a legacy process, he warns, "just gets you to the wrong destination faster".

Once that dual challenge of data and process is met, companies can finally access the creative payoff. The scale of this opportunity, however, is far larger than most realize. Citing AI thought leader Ethan Mollick, Gutiérrez-Montero notes that, "if we were to stop innovating in GenAI today, it would still take us ten to twenty years to figure out the things that we could do with the current technology."

• The new mandate: This vast potential is forcing a dramatic showdown between the new breed of "AI-native" challengers, whose processes were born already in the AI world, and the established incumbents who are "methodically but surely adapting their processes internally."
• Foundational resilience: The single biggest threat to these legacy players, however, isn't a rival's algorithm. "What I see as the biggest roadblock for the vision to realize is not only the data, but also the operating model of how technology gets built inside the walls."

Ultimately, this strategic and technological transformation unlocks something far more profound than business efficiency. It opens the door to augmenting human creativity itself. To survive, incumbents must fundamentally transform their IT departments from siloed service centers into product-centric engines embedded deep within human-business-hybrid knowledge bases. "Gone are the days where IT sits in the corner and we work on platform implementations," Gutiérrez-Montero declares. "Incumbents need to become more like the innovative new tech companies doing product development at scale."