Is AI slowing your business down?
Plenty of companies have now completed their AI experimentation stage. It’s the need to now commit that’s slowing them down, writes Paul Armstrong
AI has compressed execution across most parts of the business that matter, turning work that once took days into minutes and collapsing multi-step workflows into a single interface. Teams can produce more, test more and ship more at a pace that would have looked excessive not long ago. Faster tools should create faster companies, but firms are telling me they’re ‘feeling slower’. Capabilities aren’t the issue, their ability to choose is getting drastically pulled.
Adoption data reflects the tension. Uptake continues to rise across the UK economy, yet movement from experimentation into full deployment remains uneven, with many organisations struggling to move from testing to doing in a way that delivers real and ongoing impact.
A similar pattern appears at the moment where decisions need to convert into action. More than 80 per cent of organisations report delays or pauses when moving systems from pilots into production, largely because of unresolved questions around ownership, governance and accountability. Commitment is causing big issues for businesses of all sizes.
The decision tax
Greater capability produces more options, and not necessarily better decisions. Systems now generate strategies, forecasts and recommendations at scale, increasing the number of directions a business could plausibly take. Each comes with enough supporting logic to justify further discussion, and each adds weight to the actual process of choosing. Time shifts to reviewing, meetings expand, inputs multiply and decisions that would once have been made with incomplete information remain in circulation while more analysis is gathered. Organisations appear more rigorous on the surface, yet decisiveness erodes underneath as commitment gets pushed further and further out. Responsibility spreads across teams in a way that looks collaborative but functions more as avoidance, leaving no single owners.
Shadow systems form alongside the first, driven by employees using tools outside formal structures to get work done quickly. Both systems exist in the same company; one moves quickly and quietly, while the other is slow and visible. Trust between the two is limited, and leadership sees rising output and assumes progress even when much of that output never turns into committed action.
Faster execution, slower commitment
Execution speed is no longer scarce. Access to tools is spreading, capabilities are becoming standard and the gap between competitors is narrowing, which shifts advantage away from how quickly a company can act and towards how quickly it can decide what to do.
Firms that commit earlier, with clear ownership and a tolerance for uncertainty, move ahead while others continue to review options that all look viable, often for longer than the opportunity allows. Missed opportunities, delayed launches and slower responses to market shifts accumulate before showing up in performance, and by the time the pattern becomes visible it is already structural.
UK businesses face this more sharply than most. Highly regulated sectors dominate, governance structures are layered and risk tolerance is constrained by design, which means more scenarios to assess and more outputs to validate at exactly the point where commitment should happen. Execution accelerates into organisational resistance, widening the distance between what a company can do and what it is prepared to act on.
Many leadership teams are struggling and respond by adding more structure, more oversight and more attempts to control systems that are expanding faster than the processes around them. More control increases the cost of choosing rather than reducing it, turning speed gains at the execution layer into drag at the decision layer. Smart companies are setting new decision frameworks at the same time
Cutting the drag
Fixing this requires changing behaviour rather than adding more processes, which is exactly where most organisations default to. Every new layer extends the time between recognising a problem and acting on it, often without improving the quality of the decision itself. Delay rarely comes from a lack of information at this point. Delay comes from a lack of forced conclusion. Nobody is required to turn analysis into commitment, so the organisation continues to absorb inputs without resolving them.
Time boundaries matter more than most teams admit because decisions allowed to persist tend to expand, drawing in more data, more opinions and more justification long after the direction of travel is already clear. A defined window for commitment changes the dynamic by forcing a choice between action, escalation or abandonment, rather than allowing work to sit in a state of permanent review. Without that pressure, analysis becomes a substitute for judgement and the organisation mistakes activity for progress.
Ownership is where the issue becomes structural rather than cultural. Committees provide cover and create the appearance of rigour, yet shared responsibility rarely produces timely outcomes because no single person carries the consequence of delay. Assigning one accountable owner doesn’t reduce input or weaken challenge, but it does create a clear moment where discussion turns into commitment, which is the point most organisations struggle to reach. Without that clarity, work circulates between functions, each adding perspective, none taking responsibility for the outcome.
Many firms treat low-risk, reversible decisions as if they carry the same weight as high-risk, irreversible ones, pushing both through identical processes in the name of consistency. Separating the two clearly allows speed where it is safe and scrutiny where it is necessary, rather than applying caution universally and diluting its value in the process.
Most companies track output, productivity and delivery timelines in detail, yet very few track how long it takes to move from identifying an issue to committing to a course of action, even though that interval now determines how effectively the organisation gets results. Advantage won’t come from more tools or more data, but from organisations that can still choose when everything else speeds up, which remains a far rarer capability than most leadership teams are willing to admit.
Paul Armstrong is founder of emerging tech advisory, TBD Group, and its intelligence community TBD+