Our AI performance research reveals a clear trend: those pulling ahead don't necessarily have better technology. But they do have stronger foundations around it—trust in data, clarity of accountability, speed of decision-making, and the confidence to act before complete certainty arrives.
These are leadership challenges rather than technical ones. And they're exactly what we set out to explore when we launched PwC's Tech Leadership Exchange — a programme that brings together the UK's most senior technology leaders to solve the problems that need more than one perspective.
Trust the data or fall behind
Few businesses have AI fully embedded into day-to-day workflows. The data might exist. The dashboard might exist. But the organisational muscle to move quickly, confidently and repeatedly from insight to action often doesn't.
“Leaders making ground on AI right now aren't necessarily the ones with the heaviest investment in AI tech. They're the ones moving beyond AI experiments. They're focusing on creating solid process foundations with responsible business and IT owners, and are deploying coherent use-cases across these to create impacts you can see and measure. It's why IT is an easier place to start for many. It sounds simple, but few have done it well at scale.”
Warren Tucker, Partner, PwC UK
Our AI performance study reinforces this. UK businesses are half as likely as global AI leaders to be making active use of structured data to drive decisions (30% vs 60%). The problem isn't access to data but whether the right people have the mandate and backing to act on what the data is telling them.
The action: Build trust in data. The right governance, literacy, and cultural permission help organisations make informed decisions quickly. Organisations should look to empower people to make decisions, even at the cost of occasional failure. It's what separates those finding competitive advantage from those waiting for certainty.
Make the right people accountable
Most large organisations have accountability at the executive level. The challenge is that it often doesn't reach the people closest to the decisions that matter. For many organisations, accountability for AI-supported decisions sits with the technology function, but the consequences sit with the business. This disconnect is where oversight breaks down and transformation can stall.
The organisations scaling AI effectively have made accountability precise, visible, and structural: clear decision rights, cross-functional ownership, and a culture where owning a mistake is the best route to fixing it.
The action: Define decision rights at every level, not just at the top. Ensure people feel comfortable with debriefing as a learning practice rather than a way to apportion blame. The speed at which you learn from what went wrong is a better predictor of performance than whether the original call was right in the first place.
Shared conviction moves faster than shared data
Our AI performance study revealed our UK respondents to trail global AI leaders significantly on having a clear AI roadmap (60% vs 82%) and an AI vision aligned to business objectives (60% vs 79%). Without shared strategic clarity, even good data creates noise rather than speed.
The organisations moving fastest have an evident strategic purpose. Data is shared to encourage better connections and collaboration rather used as a reporting mechanism. When teams trust the same information and understand the same priorities, decisions can happen at pace.
The action: Make your data accessible to all and ensure alignment around it. Shared data creates shared knowledge, and shared knowledge creates speed. If timelines slip and decisions are being deferred, the problem is more likely a lack of trust rather than a lack of capability.
Fix the environment before you fix the team
One of the most interesting observations from our Tech Leadership Exchange launch — where former F1 driver David Coulthard joined the conversation — was that McLaren's recent Constructors' Championship wasn't won with new talent. It was actually won by changing the environment around the talent they already had.
“Before reaching for restructure, the question worth asking is whether your environment is set up for the people you already have to succeed. We regularly see the same teams, given the right conditions, deliver dramatically different outcomes.”
Richard Wyles, Partner, PwC UK
There's a direct parallel for technology leaders. You can have the best platform, the best data, and the best people, but if the environment doesn't give them confidence and clarity to act, you’ll still struggle to deliver.
The action: When performance isn't where it needs to be, look at the environment before the people. Consider whether underperformance is more about conditions than capability. Leaders who build infrastructure, confidence, and empowerment around their existing teams will unlock more value than those who routinely restructure.
Know what to stop—and do it quickly
UK businesses show relative strength in willingness to review AI initiatives. But willingness without clear criteria is just spinning wheels. Global AI leaders are 1.5 times more likely to use AI to create entirely new business models, not because they run more experiments, but because they're ruthless about which ones to back and which to stop.
Creating organisational permission to stop before sunk cost becomes an anchor is an important leadership skill. It requires the same trust in data and clarity of purpose that underpin everything else.
The action: Apply consistent evaluation criteria and move resource decisively when the evidence warrants it. The competitive advantage isn't in having more AI initiatives but in knowing which to back, which to cut, and having the discipline to do both before your competitors.
Five actions to create one compounding advantage
These aren't separate challenges. They're interconnected:
- Trust in data enables faster decisions
- Clear accountability makes learning possible
- Shared purpose aligns speed with direction
- The right environment unlocks existing capability
- The discipline to stop frees up capacity to focus
Organisations that build all five will create a compounding advantage. Those missing even one will continue to wonder why their technology isn't delivering what it promised.
