Over the past year, I have watched organisations shift from asking whether they should adopt AI to working out how to manage it.
Spending continues to rise, even while many businesses are still debating long-term return on investment. This suggests the next phase of AI adoption will be shaped less by the technology itself and more by governance, visibility and accountability.
AI spending has entered a new phase
Earlier this year, Forbes reported that Uber exhausted its entire annual AI budget in just four months, prompting the company to introduce internal caps on tools such as Claude Code and Cursor. While the numbers are staggering at Uber’s scale, they also show how quickly AI spending can grow as adoption accelerates.
According to reports, the surge in spending was driven almost entirely by the rapid adoption of AI coding platforms across engineering teams. There was no failed transformation programme, major technology mistake or obvious misuse. Developers simply found the tools useful, adoption accelerated, and costs increased alongside it.
Uber’s experience reflects a broader trend. For the past two years, much of the conversation around AI has focused on productivity. Increasingly, that conversation is being matched by spending patterns, as organisations continue to increase investment as adoption grows.
Soldo’s latest Spending Trends Index found that average corporate spending on AI tools increased by 77% in 2025, while investment across the ten most popular AI platforms grew by 175%.
Alongside that growth, however, sits a question that occupies boardrooms, CEOs, and CFOs alike: how much measurable business value is actually being created?
Productivity is no longer the biggest question
Many organisations can point to examples of improved productivity. Developers write code faster and teams spend less time on repetitive tasks. Connecting those improvements to revenue growth, profitability or competitive advantage is much harder.
That uncertainty is exactly what makes the rise in AI spending so interesting. The decision to adopt AI is becoming less of a debate. The bigger question now is how organisations govern it.
Governance is becoming the competitive advantage
Governance is hardly the most fashionable word in business. It lacks the appeal of innovation or disruption, yet it ultimately determines how organisations extract value from technology. It defines who can access tools, who approves spending, how costs are monitored and how return on investment is measured.
These questions are becoming increasingly urgent because software purchasing has transformed.
Only a few years ago, technology adoption followed a relatively predictable process. A team identified a need, procurement stepped in, budgets were approved and finance retained visibility throughout the purchasing journey.
AI is changing how software gets bought
Today, software purchasing is far more decentralised. A marketing team can subscribe to ChatGPT in minutes, an engineer can adopt Cursor before anyone outside the team knows it exists, and product managers can run pilots without waiting for an annual budgeting cycle.
Individually, none of these purchases are likely to concern a CFO. Collectively, however, they become a significant category of spend.
Anyone who has managed software budgets over the past decade will recognise this trend. AI, however, accelerates it by introducing an extra layer of complexity. Costs are rarely fixed. The more valuable an AI tool becomes, the more a team uses it, meaning success and expenditure scale together.
Why visibility matters more than ever
Recent research from Pricing I/O and Benchmarkit illustrates this clearly, finding that almost nine in ten organisations have exceeded their original AI budgets. Crucially, very few attributed this to vendor price increases. Instead, they had simply underestimated how quickly internal adoption would grow. This is fundamentally a governance challenge rather than a pricing one.
Without real-time visibility into which tools are being used, who is purchasing them and how spending is evolving alongside business outcomes, it becomes difficult to evaluate true value. After all, it is incredibly hard to measure a return on investment when you are still trying to map the investment itself.
These are conversations we are increasingly having with finance leaders. While discussions have traditionally focused on travel, procurement and software subscriptions, AI is quickly becoming a central topic.
Finance is not trying to slow AI down
Finance leaders are not looking to slow innovation. In many cases, they are actively encouraging experimentation. What they need is visibility: understanding what is being purchased, how spending is evolving and whether investment is delivering measurable business outcomes.
Traditional reporting cycles were built for a much slower pace of decision-making. When an AI tool can be purchased in minutes and embedded into everyday workflows within weeks, discovering its cost at the end of the month often comes too late. By then, adoption has already accelerated and spending has become embedded.
Good governance is sometimes viewed as a constraint on innovation, but effective guardrails create confidence. They allow organisations to experiment more freely because leadership understands what is happening.
The next phase of AI will be defined by governance
AI will not be the last technology to challenge traditional governance models, but it is undoubtedly one of the fastest-moving.
As adoption continues to scale, the organisations that benefit most won’t just be those with the largest AI budgets. The real advantage goes to those tracking how it’s used, what it costs, and the value it creates.
As AI adoption scales, the organisations that benefit most won’t be those with the largest budgets, but those with the clearest understanding of how AI is being used, what it costs and the value it creates.
Governance isn’t there to stop AI adoption. It’s what gives organisations the confidence to scale it.
