How to beat the AI productivity trap: Or, why working faster isn’t working smarter

There is a paradox at the heart of enterprise AI adoption, and finance leaders are caught squarely in the middle of it.

Ask employees whether AI has made them more productive and the answer is overwhelmingly yes. Soldo’s 2026 Productivity at Work report found that 88% of employees say AI tools have improved their productivity.

But ask CEOs whether that productivity is translating into financial returns, and the picture looks very different. PwC’s 29th Global CEO Survey, based on responses from 4,454 chief executives across 95 countries, found that 56% say AI has delivered neither revenue nor cost benefits to date. Only 12% say it has delivered both.

Speed, it turns out, is not the same as value.

The measurement gap

The disconnect is not just statistical noise. It points to a structural problem in how organisations are thinking about AI. Many are measuring the wrong things. Employees are faster. Tasks take less time. Reports get drafted in minutes rather than hours. These are real gains, but they are activity metrics, not business outcomes. The board wants to know what the AI investment is actually worth. Most organisations still can’t tell them.

A CloudZero survey of 260 finance executives found that 87% say they need to tie AI spend to business outcomes, but only 22% can actually do so right now. Meanwhile IDC’s 2025 AI Investment Survey found that 61% of enterprises can’t demonstrate measurable ROI from their AI investments because they never established a baseline before deploying.

This is the trap.

Businesses are moving fast without a measuring stick, then expressing surprise when they can’t show the results.

Finance is losing control of the spending

The governance problem compounds this measurement problem. Unlike a traditional technology purchase – a single licence, a defined rollout, a budget line – AI adoption is sprawling. According to the PWC survey, it happens through subscriptions, usage-based pricing and departmental credit cards, accumulating across teams in ways that no single function is tracking.

Soldo’s research found that only 27% of finance leaders say clear policies and controls fully govern AI use and purchasing. The rest are operating in varying degrees of the dark: uncertain about which tools are in use, who authorised them, what data is being shared, and what the total bill actually is.

When processes are opaque and slow, workarounds follow. Six in ten employees say they frequently bend rules or find loopholes to access spending they need.

It’s worth reflecting on what that really means: the majority of workforces have found it easier to circumvent finance controls than to work within them.

As IBM’s Monica Proothi, Global Finance Transformation Lead, put it: “It’s not AI that will take your job – it will be the person who knows how to use it.”

The same logic applies at an organisational level. It isn’t AI that will outpace your business but a competitor that has figured out how to govern it.

Productivity at Work – Soldo 2026 Survey Report

This report explores the challenges businesses face in aligning attitudes and approaches to productivity. How can finance leaders overcome barriers and position themselves as drivers of growth? Download to learn more.

The finance function’s moment

This is where the CFO’s role becomes critical. And arguably more interesting than it has ever been.

We all know that finance isn’t just a reporting function any more. As we explored in our coverage of how enterprise architecture platforms are becoming the hidden foundation of AI governance, the connective tissue between AI adoption and business value is governance infrastructure.

Finance sits at the centre of that infrastructure.

According to Soldo’s research, almost three quarters of finance leaders say the finance function is now responsible for overseeing AI investment and performance. Employees themselves see it this way too: 41% describe finance as an enabler. That’s a significant shift from the traditional perception of finance as a gatekeeper.

The businesses pulling ahead on AI ROI aren’t necessarily the ones with the most sophisticated models or the largest datasets. They are the ones that embedded governance before they scaled, connected spending to outcomes, and gave finance teams real-time visibility to make decisions rather than write retrospective reports.

Governing the paradox

Soldo’s Brandon Till, Head of Transformation, frames it cleanly: productivity gains should be achieved by actively supporting workers on their AI journey, helping them build skills, removing friction, and enabling them to contribute at a higher level. That is not a technology problem. It is a process and governance problem.

The fix starts with visibility. What AI tools is the organisation actually using? How much do they cost, who owns them, what are they delivering in terms of business value? That’s the baseline for any meaningful ROI conversation.

From there, finance leaders can set the terms of engagement, approving tools that demonstrate value, cutting those that don’t, and building the measurement frameworks that connect AI activity to the financial outcomes boards actually care about.

Until finance teams can draw a straight line between the means – workers saying they feel more productive thanks to AI – and the end of a business that runs better, costs less and competes more effectively, the productivity paradox will persist. And the AI budget will remain, for many organisations, an act of faith rather than an investment.