By Ankit Sarawagi

Capital allocation has always reflected what a business believes will matter most over time. What has changed over the last few years is not just the size of these investment decisions, but how uncertain they have become. AI has added another layer to what was already a complex equation, and it is still taking shape. In many cases, there isn’t a clear line of sight on how these investments will play out, or how long they will take to show real returns, which makes planning around them a bit more tentative than it used to be.

Technology, talent, and growth also don’t sit as neatly as they once did. They are still talked about as separate priorities, but in practice they tend to pull from the same pool of capital. That is where things start to get tricky. Each of them moves at a different pace, and that difference matters more now. Some investments show results fairly quickly, others take longer to settle, and it is usually in that gap that the pressure starts to build.

AI investment has ramped up sharply over the past year. In 2025 alone, global spending crossed $225 billion, with a growing share of venture funding moving toward AI-led businesses. Large technology companies are expected to commit over $650 billion toward AI infrastructure by 2026, which starts to put the scale into perspective. This doesn’t feel like a routine increase in spending. It points to a broader shift in how AI is being treated within businesses.

Technology as an Ongoing Investment Layer

For most organisations, the question is no longer whether to invest in AI, but how to do it without throwing the rest of the business out of balance. It sounds simple enough when you put it that way, but it tends to get complicated fairly quickly once you start working through what that actually involves.

Unlike traditional technology investments, this is not a one-time build followed by a clear upgrade cycle. It tends to keep expanding. What begins as a focused investment in one area often leads to additional spend across data systems, compute requirements, and integration work that was not fully visible at the start. Over time, it becomes less of a project and more of an ongoing layer that the business has to keep supporting.

Returns also behave differently. They are often uneven—showing up in parts or over longer periods—which makes them harder to track against a single metric. This mismatch between ongoing investment and staggered returns is what increasingly shapes capital allocation decisions.

How AI Is Reshaping Talent Decisions

The impact of AI on talent is already visible, although it does not always show up in obvious ways. In day-to-day operations, AI agents are not just reducing work, they are also changing what that work looks like. How that plays out depends a lot on how organisations choose to use them, and that tends to vary more than expected.

In setups where the implementation is more considered, repetitive and high-volume tasks gradually move toward automation. That shift is not always immediate, but it becomes noticeable over time. As that happens, human roles begin to tilt toward areas that are harder to standardise, such as judgment, relationship management, and dealing with more complex or ambiguous situations where context matters. At Verloop, this shift is evident in customer support. As AI agents take on more routine interactions, the need for large frontline teams starts to come down. The shift isn’t uniform, though. In many cases, demand moves toward roles that deal with escalations, quality checks, and the overall customer experience. The work itself doesn’t go away. It just moves to areas where the stakes, and the value, are higher.

When AI is used primarily as a cost-cutting tool, the results don’t always hold up over time. Teams may become leaner, but some of the trade-offs are easy to miss in the early stages, and they don’t always show up immediately. Over time, though, the impact becomes more visible.

Depth can start to thin out, and that tends to affect execution in ways that are not always obvious at first. It is usually in areas that rely on experience or context where this shows up more clearly, which is why the initial efficiency gains don’t always translate into sustained outcomes. In some cases, the immediate efficiency gains end up masking these gaps until they become harder to address.

Organisations that are seeing more sustainable outcomes tend to approach AI a little differently. Instead of using it primarily to reduce headcount, they use it to support and extend what teams can do. That changes the way roles evolve and how work gets distributed, but it also tends to preserve a certain level of depth that is easy to lose otherwise. The technology itself does not really dictate the result; most of it comes down to how it is introduced and used within the organisation, which is why outcomes vary as much as they do.

Growth Under a Different Lens

Growth, which has traditionally absorbed the largest share of capital, is also being reconsidered. For a long time, the assumption was that growth could be accelerated as long as capital was available, with efficiency catching up later. That model is under pressure. AI has made it easier to scale certain functions, but it has also introduced new cost layers, particularly around infrastructure and deployment. The result is that growth can appear strong at the surface while underlying efficiency is still catching up, which makes capital allocation more sensitive to how that growth is being funded.

Why These Decisions Are Becoming Harder to Separate

These decisions are more connected than they look at first. The kind of technology a company invests in ends up shaping how teams are structured, sometimes in ways that aren’t obvious right away. And even then, the outcome still depends on the people using it. The same tool can produce very different results depending on the team behind it. Growth usually follows from that mix, not from any one piece in isolation.

That’s also where things get complicated. It’s still tempting to treat technology, talent, and growth as separate priorities, at least for a while. But that separation doesn’t really hold for long. Misalignment tends to show up later, and by then it is often more difficult, and more expensive, to fix. The real challenge now is less about choosing one over the other, and more about keeping them in step as the business evolves.

Discipline, Not Access, Is the Differentiator

Access to capital is no longer the primary differentiator. Discipline is. AI has expanded what companies can build, but it has also increased the cost of getting decisions wrong. Investments are larger, timelines are longer, and course correction is more difficult once capital is committed.

Organisations navigating this well tend to apply the same financial discipline to AI as they do elsewhere—defining outcomes early, linking investments to measurable signals, and adjusting when those signals do not appear.

In that sense, capital allocation today is less about choosing between technology, talent, and growth, and more about understanding how they work together. The companies that get this right are not always the ones investing the most, but the ones that are clearer about where returns will come from and how long they will take.

The author is CFO, Verloop.io

Disclaimer: The views expressed are the author’s own and do not reflect the official policy or position of Financial Express.