Is the Traditional IT Team Structure Still Fit for Purpose in 2026?

By now, most technology leaders have sat through enough AI presentations to last a lifetime. 

We've heard that AI will transform productivity. We've heard it will replace tech jobs. We've heard it will revolutionise software development, service management, project delivery and just about every other function within an organisation. 

Some of that will undoubtedly prove true. But I think we're asking the wrong question. 

The conversation shouldn't be about whether AI will change technology teams. That's already happening. 

The more important question is whether the traditional technology team structure was fit for purpose before AI arrived. 

Because when we speak with CIOs, CTOs and technology leaders, the frustrations they describe aren't new. 

Projects take too long to deliver. Teams work in silos. Priorities change faster than delivery roadmaps can keep up. Technology functions struggle to recruit specialist skills. Business stakeholders become frustrated by complexity, while technology teams become frustrated by unrealistic expectations. 

None of those problems were caused by AI. 

Many organisations were already wrestling with them long before ChatGPT entered the mainstream. 

What AI is doing, however, is exposing those weaknesses far more quickly than before. 

AI Isn't Breaking Technology Teams. It's Exposing Their Weaknesses. 

For years, organisations have structured technology teams around specialist disciplines. 

Development. Infrastructure. Security. Data. Testing. Business Analysis. Project Management. 

That approach made perfect sense when technical execution was the primary challenge. 

Work flowed from one specialist team to another. Requirements were gathered, solutions designed, code developed, systems tested and projects delivered. 

The model wasn't perfect, but it was logical. 

The problem is that many of those activities are now being accelerated by AI. 

A developer can generate boilerplate code in seconds. Testing tools can create test cases automatically. Documentation can be produced faster than ever before. AI-powered assistants can help analyse requirements, summarise meetings and surface insights from vast quantities of information. 

As a result, the bottleneck is beginning to shift. 

In many organisations, technology is no longer the thing slowing delivery down. 

Decision-making is. Alignment is. Governance is. Leadership is. 

The uncomfortable reality for many businesses is that AI may improve productivity, but it won't fix organisational dysfunction. 

If anything, it highlights it. 

You May Still Be Hiring for Yesterday's Problems 

One of the most interesting conversations happening within technology leadership circles right now is not about AI tools. It's about workforce planning. 

Specifically, whether organisations are continuing to recruit for team structures that are gradually losing relevance. 

That's not to suggest roles are disappearing overnight. 

Far from it. 

Business Analysts, Test Managers, Project Managers, Data Engineers and Technical Specialists continue to play important roles within most organisations. 

But leaders should be asking tougher questions. 

Are these roles delivering value because of their expertise? 

Or because existing processes require them to exist? 

There is a significant difference. 

Historically, many organisations created specialist functions to bridge gaps between systems, teams and processes. 

Today, AI is beginning to remove some of that friction. 

The challenge is that workforce planning often lags behind technological change. 

Many organisations are still scaling teams based on assumptions that were formed five or ten years ago. 

Meanwhile, the work itself is evolving. 

The future may not require fewer technology professionals. 

It may require professionals with broader skillsets who can operate across multiple domains, supported by AI rather than constrained by traditional functional boundaries. 

The Productivity Question Nobody Can Answer Yet 

Perhaps the most overused claim in technology today is that AI will dramatically improve productivity. 

It probably will. The problem is that nobody seems entirely sure what happens next. 

If a software engineer becomes twice as productive, does the organisation need half as many engineers? 

Or does it deliver twice as much value? 

History would suggest the latter. 

Technology leaders have spent decades dealing with growing demand for digital services, applications, integrations, security requirements and data capabilities. 

Every time technology becomes more efficient, organisations tend to find new opportunities to create value. 

Cloud computing didn't reduce demand for technology talent. Neither did automation. Neither did low-code platforms. Instead, they shifted where effort was spent. 

There is a strong argument that AI will follow a similar pattern. 

The organisations that benefit most may not be those reducing headcount. They may be those using productivity gains to accelerate innovation, modernisation and business transformation. 

There is also a growing gap between AI adoption and measurable business outcomes. According to McKinsey's latest State of AI research, 64% of organisations say AI is driving innovation, yet only 39% report meaningful impact on profitability. That's a significant disconnect. It suggests many organisations are still experimenting, piloting and learning rather than fundamentally transforming how they operate. 

The Real Skills Gap May Be About Judgement, Not Technology 

For all the discussion around AI skills, I think many organisations are focusing on the wrong capability gap. 

The assumption is often that people need to learn how to use AI tools. 

That is certainly part of the equation. But the more important capability may be knowing when not to trust them. 

As AI-generated outputs become increasingly sophisticated, the value of critical thinking, commercial awareness and professional judgement rises significantly. 

Anyone can generate code. Anyone can ask AI to produce a business case. Anyone can create a project plan. 

The differentiator is understanding whether the output is any good. Technology leaders are increasingly telling me the same thing. The challenge isn't getting people to use AI. 

It's ensuring they understand its limitations. That requires a different type of workforce development strategy. 

One focused less on technical training and more on decision-making, problem-solving, communication and business understanding. 

The World Economic Forum's Future of Jobs Report found that employers expect 39% of workers' core skills to change by 2030. That's not simply a technology challenge. It's a leadership challenge. Organisations need to think carefully about how they develop commercial judgement, critical thinking, communication and decision-making alongside technical capability. 

Ironically, the more capable AI becomes, the more valuable these human skills are likely to become. 

What If AI Adoption Doesn't Happen as Fast as Predicted? 

This is perhaps the question that receives the least attention. 

Most discussions about the future of technology teams assume AI adoption will continue accelerating at its current pace. 

That may well happen. 

But history suggests technology adoption is rarely a straight line. 

The reality inside many organisations is more complicated. 

Security concerns remain. Governance frameworks are still developing. Regulatory expectations continue to evolve. 

Questions around intellectual property, data privacy and accountability remain unresolved in many industries. 

For sectors such as healthcare, financial services and government, large-scale adoption may take considerably longer than many forecasts suggest. 

There's also the issue of return on investment. 

Many organisations have enthusiastically adopted AI tools because they fear being left behind. Far fewer have reached the stage where they can confidently demonstrate measurable business outcomes. 

That doesn't mean AI won't deliver value. It means we're still relatively early in the journey. 

That's important because workforce planning decisions made today are often based on assumptions about where technology will be in three to five years. 

If those assumptions prove incorrect, organisations risk redesigning teams for a future that arrives more slowly than expected. 

The winners are unlikely to be those who make the most aggressive bets. They will be those who remain adaptable. 

What Might Technology Teams Look Like By 2030? 

While nobody can say with certainty what technology teams will look like in five years, there are some clear trends beginning to emerge. 

Smaller, multidisciplinary teams 

Rather than large departments separated by function, teams may increasingly be built around products, customer outcomes or business problems. 

Greater emphasis on adaptability 

Technology professionals may be expected to work across multiple disciplines, supported by AI tools that reduce the need for narrow specialisation. 

New leadership responsibilities 

Managing people will increasingly mean managing a combination of human capability and AI-enabled workflows. Leaders will need to understand productivity, governance and risk from both perspectives. 

Different hiring priorities 

The most sought-after technology talent may not necessarily be the most technically specialised. Organisations are likely to place greater value on critical thinking, communication, commercial awareness and the ability to adapt alongside rapidly changing technology. 

This is already influencing technology recruitment across New Zealand, with employers increasingly prioritising adaptability, learning agility and business understanding alongside technical expertise. 

What Should Technology Leaders Be Doing? 

Technology leaders should spend less time asking how AI fits into their existing structures and more time questioning whether those structures still make sense. 

• Where does work actually get stuck?
• Where are handovers creating unnecessary friction?
• Which roles exist because they add value, and which exist because of process complexity?
• What skills will become more important as AI becomes more capable?
• How should leadership evolve when technical execution becomes easier but organisational complexity remains? 

These are not easy questions. 

But they are probably more important than deciding which AI platform to deploy next. 

Because technology has never really been the hard part. 

The hard part has always been helping organisations change. 

The Bigger Risk Isn't AI 

AI is forcing organisations to confront questions they have been avoiding for years. 

Questions about team structures, skills, leadership, and whether existing operating models are genuinely designed for the pace and complexity of modern business. 

The organisations that gain the most from AI won't necessarily be those investing the most money in AI platforms. 

They'll be the organisations willing to challenge long-held assumptions about how technology teams should be built in the first place. 

Because the biggest risk facing technology leaders in 2026 isn't that AI replaces jobs. 

It's that businesses spend the next five years applying AI to operating models that were already struggling to keep up.