Expert answer / 19
Are we ready to spend on AI?
Short answer
Start with a business problem, an accountable owner and a safe way to measure whether AI improves the work.
More detail
Questions behind the answer
Whatis happening
What does being ready for AI mean?
Being ready for AI means having a worthwhile use case, an owner, suitable data, a workable process, trained people and clear safeguards.
Whyit happens
Why do AI projects stall after the trial?
AI trials stall when they are separate from real workflows, have no success measure or leave teams unsure about accuracy, privacy and accountability.
Howto fix it
How do we choose our first AI project?
- Choose a repeated task with a clear time, cost or service problem.
- Check data access, privacy, accuracy risks and who reviews the output.
- Run a bounded pilot against a baseline and agree in advance what would justify scaling.
Sources: The State of AI in 2025: Agents, Innovation, and Transformation - McKinsey · From Adoption to Impact: Three Horizons of AI Transformation - McKinsey