Getting AI to work isn't an IT problem, an HR problem or a strategy problem. It's all three at once. Yet most organisations give ownership to one function and leave gaps nobody addresses. IT focuses on infrastructure and software. HR focuses on people. Strategy focuses on business outcomes.
The work in between, including organisational memory, codifying decisions and redesigning work, falls through the cracks. That isn't a failure of any one function. It's a coordination challenge that needs a new organisational muscle. The next decade belongs to organisations that learn to bridge these silos and treat AI deployment as a shared, interdisciplinary capability.
Whose job is AI deployment?
When organisations decide to "get ready for AI", the first question is usually: whose job is this?
IT says it's about infrastructure and software: data pipelines, model deployment, system integration. That's right, but it isn't the whole picture.
HR says it's about people: skills development, change management, workforce planning. Also right, and also incomplete.
Strategy says it's about business outcomes: competitive advantage, operational efficiency, new capabilities. Correct again, and still incomplete.
Why it needs all three functions at once
The problem is that deploying AI well needs all three at the same time, and most organisations aren't set up to work that way.
Consider what it actually takes. You need technical groundwork: data that's clean enough, integrated systems, reliable models. That's IT's domain. You need organisational capability: people who can work alongside AI, and processes designed for human-machine collaboration. That's HR territory. You need strategic clarity: which outcomes matter, where AI adds value, how to measure success. That's strategy's responsibility.
The work that falls between functions
The critical work sits between these functions. Codifying organisational decision logic so AI can use it. Designing workflows that split tasks between people and machines. Building data architecture that connects people, work and technology. Setting up governance for AI safety and ethics while still allowing fast experimentation.
This work doesn't naturally belong to IT, HR or strategy. It needs all three working together with shared accountability.
What happens when one function owns it
Most organisations default to giving ownership to one function. IT gets it because AI sounds technical. Or HR, because it's about the workforce. Or strategy, because it's transformation.
The result is critical gaps. IT builds infrastructure without understanding how work actually flows. HR designs training without knowing what the technology can reliably do. Strategy sets direction without seeing the constraints on the ground.
Nobody is failing at their job. Everyone is working within their own expertise. The problem is that AI deployment crosses traditional boundaries.
How the organisations making progress work
The organisations making progress treat it as an interdisciplinary capability. They form cross-functional teams with joint accountability. They create roles that bridge IT, HR and strategy, connecting all three rather than reporting into one. They set up governance where technical, organisational and strategic perspectives carry equal weight.
This isn't matrix management or another coordinating committee. It's rethinking who owns what when the challenge doesn't fit existing boxes.
The next decade will separate the organisations that work this out from those that don't. Not because one group has better AI or smarter people, but because some will build the muscle to own interdisciplinary challenges while others keep optimising within functional silos.
AI is just the first test. The real challenge is building organisations that can tackle complex problems with IT, HR and strategy working as one system, not three functions coordinating at the edges. That capability, more than any single AI implementation, decides who wins.
