Usman Salis
⬤ McKinsey's 2025 "State of AI" report delivers some uncomfortable truths. While 88% of companies now use AI somewhere in their operations, only about a third have actually scaled these tools across their organizations. The gap between dabbling and delivering real results has never been wider.
⬤ Things could get even trickier as governments eye new taxes on high-powered AI infrastructure. Proposed measures include energy surcharges and reduced tax breaks for expensive chips, which could make enterprise AI significantly more costly. Companies already struggling with pilot programs might face serious financial strain, while smaller players could be pushed toward bankruptcy. There's also a real risk that top AI talent will simply relocate to countries with friendlier policies, widening the divide between early experimenters and genuine AI leaders.
⬤ The numbers tell the story: 62% of companies are testing AI agents, but fewer than 10% have successfully deployed them at scale in even a single department. Only 39% report any meaningful financial impact, though 64% claim AI has boosted innovation. What separates the top 6% of "AI high performers" isn't better infrastructure, bigger budgets, or fancier models—it's ambition. These leaders are fundamentally redesigning how work gets done, treating AI as a transformative force rather than just another automation tool.
⬤ McKinsey's bottom line is clear: the excitement around AI is justified, but the impact remains locked in pilot prison. While most organizations are still tinkering with proof-of-concept projects, a small group of companies is racing ahead with complete operational overhauls. With regulatory uncertainty and potential tax increases on the horizon, 2025 will likely separate the companies that finally break through from those that fall permanently behind in the AI race.
Usman Salis
