The Case for Slowing Down on AI

The advice is everywhere: move faster on AI. Adopt more. Automate more. Deploy more. It comes from consultants, from technology vendors, from the business press. And most leaders are nodding along, feeling the pressure to keep up. I think they are missing something.

Speed without direction is expensive chaos. And right now, a significant proportion of organisations are discovering that the hard way.

What the data actually shows

BCG's AI research from September 2025 found that 35% of organisations are scaling AI faster than they feel comfortable with - their own admission that they lack the foundational readiness to support the pace. The same research found that the majority of businesses report minimal revenue or cost gains from adoption. That is not a technology problem. That is a direction problem.

The organisations BCG classifies as "future-built" - the 5% who plan to spend more than twice as much on AI and expect twice the revenue increase - are not simply moving faster. They are moving with clarity. The research distinguishes between reactive urgency and strategically-aligned adoption. It is a meaningful distinction. One produces activity. The other produces advantage.

The flaw in the "move fast" consensus

The consensus view assumes that the risk of moving too slowly outweighs the risk of moving without purpose. I think that assumption is backwards.

What it misses is that AI does not deliver value on its own. London Business School's research from April 2026 makes this point clearly: AI success depends less on technology selection and more on leadership design - specifically, how AI is embedded in real workflows. The critical variable is not which models or vendors you choose. It is the quality of the human architecture around the tool.

When leaders rush that architecture, they skip the decisions that actually determine whether AI works: which workflows to target, what roles need to shift, what safeguards matter in domains where errors are consequential. Leaders who insist on immediate ROI from AI, the same LBS research found, risk stalling the learning cycle entirely. Frantic adoption does not avoid this trap. It deepens it.

How should business leaders think about AI strategy?

A strategic pause is not a delay - it is the deliberate act of stepping back to establish direction before accelerating. The leaders who will navigate this period best are not those who move fastest. They are those who pause long enough to work out where they are going. Organisations that establish clear leadership design and embed AI with intention will compound their advantage faster than those who accumulate disconnected tools under pressure. The pause creates the conditions for the speed to matter.

What the pause actually requires

The human architecture AI needs is not complicated to describe, but it does require time to build. It requires leaders who know which decisions AI should inform and which it should not touch. It requires psychological safety - Korn Ferry's January 2026 research identifies emotional intelligence as essential infrastructure for AI adoption, not a soft skill. When employees anticipate loss rather than opportunity, the technology meets fear-based resistance, and that resistance compounds.

Harvard Business Review's analysis from March 2026 frames it well: as AI absorbs more execution-oriented tasks, human roles must shift toward design, orchestration, and interpretation. Without that explicit redesign, AI produces activity rather than advantage. You cannot design those roles well under pressure. That is exactly the kind of thinking that gets skipped when organisations move faster than they are ready to.

The soft close

I could be wrong about this. The pressure to move is real, and sitting still is not the answer. But the leaders I find most interesting right now are not asking "how do we go faster?" They are asking "what does good look like, and how do we build toward it?"

That is a different question. It is also, I think, the more commercially valuable one. Speed matters - but only once you know where you are pointed. The organisations that slow down first will compound fastest. That is not a wellness argument. It is a strategic one.

If you want to work through what that looks like in your organisation, the AI Leaders Fellowship is where senior leaders are building exactly that kind of direction - not a collection of AI tools, but a strategy worth accelerating.