Gen AI is everywhere in companies

The gen AI paradox

Most organisations have now “done something” with gen AI. Internal chatbots, document copilots, meeting summaries and email helpers are common. These horizontal tools are easy to turn on and make individual workdays a bit smoother. But the gains are thinly spread and hard to measure. They rarely change how a business actually runs.

The real bottom-line impact comes from deeper, vertical use cases: AI embedded into specific processes like claims, pricing, forecasting, recruitment, customer operations, or compliance. That’s where time, errors, and cost come out—and where revenue and margin can move.

Why AI efforts stall

A few patterns keep showing up:

• Lots of pilots, not much scale. Teams try gen AI in pockets, but nothing joins up and nothing is fully owned.
• Tools without redesign. AI gets added on top of old workflows instead of being used to rethink how work is done.
• Not enough of the right skills to connect AI to outcomes, governance, and change.

Without people who understand both AI and the business, initiatives stay stuck at the “cool demo” stage.

What actually shifts the P&L

The organisations that do see impact tend to:

• Start with a clear business problem (cost, risk, revenue), not a technology wish list.
• Choose a narrow, high‑value process and redesign it with AI built in, not bolted on.
• Line up ownership, data, and change management from day one.
• Build something that can run after the pilot—simple metrics, guardrails, and clear roles.

In other words, the next wave of value won’t come from more copilots—it will come from fewer, better‑designed, function‑specific AI solutions that are wired into how work actually gets done.

‍