Skip to content
Fast Company Executive Board

The Fast Company Executive Board is a private, fee-based network of influential leaders, experts, executives, and entrepreneurs who share their insights with our audience.

The AI-first company requires a work redesign

Why your company hasn't changed despite an AI launch.

[Photo: Getty Images]

Profile photo of Guillaume Lardeux

Companies spent the past two years deploying AI copilots, experimenting with assistants, and automating tasks. Despite some pockets of productivity gains, a common question remains in boardrooms and leadership meetings: Why hasn’t all that AI fundamentally changed how the business operates?

Because most organizations have focused on adding AI to existing ways of working rather than redesigning work around AI.

AI’s next transformational phase requires a different discipline. Leaders must ask how to redesign work when they embed intelligence directly into it.

I’ve seen this pattern across multiple technology transitions. From on-premises systems to cloud, from cloud to digital, and now from digital to AI, the companies that created lasting advantage changed how the business worked around it.

AI-enabled vs. AI-first

A recent World Economic Forum report reinforces this shift. The organizations creating the most value from AI are redesigning workflows, roles, and who makes which calls. Intelligence becomes part of how the enterprise runs, not simply layering new tools onto existing processes. That is the difference between being AI-enabled and becoming AI-first.

This aligns closely with what we are learning through our own AI-first transformation journey. We continue to examine how work moves across the company, identifying workflows with the most friction, complexity, and cost when teams hand off to each other. Our goal is to embed AI into how we operate and redesign work to improve business outcomes.

A good example comes from our product support organization. Support engineers spent significant time on administrative work before they could focus on solving complex technical issues.

We redesigned the end-to-end support workflow and embedded AI across key stages to automate repetitive work, surface relevant context, and guide next steps. Support engineers remained responsible for technical diagnosis, customer communication, and complex decision-making.

The results have included more than 157,000 cumulative working hours saved, a 20-point increase in Customer Experience scores, and an average of five minutes saved per engagement.

Start with the workflows that hurt most

Not every workflow deserves the same level of AI investment. Leadership must choose where to concentrate intelligence. The strongest opportunities often sit in workflows with high volume, high friction, or the hardest judgment calls.

In practice, becoming AI-first means redefining high-value workflows end to end. Leaders must determine where AI should act autonomously, where it should assist people and where human judgment must remain in the lead.

This focus on workflows increasingly separates leaders from laggards, and the most valuable transformation efforts target how work is executed rather than where you can deploy AI.

McKinsey’s The State of AI research identified AI “high performers,” or respondents who attribute “significant value” and more than 5% of earnings before interest and taxes (EBIT) to AI. These higher performers are three times more likely to fundamentally redesign workflows as part of their AI initiatives.

Yet this is where many organizations get stuck. They add AI to legacy systems and existing processes, then expect a dramatic jump. But if the workflow still depends on the same handoffs, unclear ownership, and disconnected data, AI becomes just another layer on top of existing complexity. The operating model has to change for the technology to reach its potential.

From copilots to agents

This becomes even more important as companies move from copilots to agentic AI. An AI agent handling IT operations, customer service, underwriting, or supply chain coordination cannot be effective as a disconnected tool. The agent needs to share context, access the right systems, and have clear permissions, escalation paths, and governance boundaries, within an operating model built to coordinate all of this company-wide.

That elevates the role of people, as human contribution becomes increasingly concentrated around judgment, creativity, and accountability. A leader’s role is to intentionally design how those capabilities work together within critical workflows, so humans and AI complement one another rather than operate in parallel.

Customer experience can offer a clear example of this redesign already in action. Customers experience the outcomes of a service issue involving sales, support, operations, and technology teams across multiple systems.

Consider a customer reporting a suspicious transaction. Today, data from fraud systems, transaction history, customer records, and prior interactions must move seamlessly across multiple teams to deliver the right outcome.

An AI-first operating model ensures that context follows the workflow rather than being trapped in individual systems or handoffs. As AI becomes more autonomous, it can act within that workflow—gathering relevant information, identifying likely fraud, initiating the appropriate investigation, and escalating only decisions that require human judgment. The result is a faster, more coordinated experience for the customer, while employees focus on the moments where expertise and accountability matter most.

Where to start

For business leaders, this is a practical call to action. Pick the workflows that matter most to customers, employees, and business performance. Redesign them around outcomes. Decide where AI should act, where people should lead, and who stays accountable. Then build the governance, data, and measurement needed to scale with confidence.

AI-first transformation requires a deliberate operating model that embeds intelligence into how decisions are made, work is coordinated, and value is created. Pilots prove what AI can do. Operating models determine what value it ultimately creates. The companies that get this right will be defined by how effectively they redesign the work around it.

Guillaume Lardeux is the head of transformation and office of the CEO at Genesys.