Once upon a time, if your computer stopped working, you called IT.
Forgotten your password? Call IT.
Printer refusing to print? Call IT.
Clicked on something you probably shouldn’t have? Definitely call IT.
For decades, the IT department has been responsible for keeping workplace technology running while the rest of us remain blissfully unaware of what happens behind the scenes.
But as AI becomes embedded in everyday work, organisations are going to need a new department for their organisational charts.
Meet the AI Department. No, it won’t just be IT with a chatbot.
The AI Department of the future may have a very different job.
While IT manages the technology, an AI Department would focus on how the organisation actually works with it.
Its job would not simply be to approve tools or troubleshoot why your chatbot is behaving strangely. It would sit somewhere between technology, learning and development, governance, risk and organisational improvement.
Meet your new AI colleagues
The AI Department may oversee an increasingly complicated workforce. There will still be humans, obviously who spend their days asking questions such as:
- What should humans be doing?
- What should AI be doing?
- Where should they work together?
- And who is checking that any of this is actually working?
But alongside them may sit AI assistants, specialised agents and automated workflows quietly completing tasks in the background. The AI guys will need to know:
- What these systems are doing?
- Who authorised them?
- What information can they access?
- Which decisions can they make?
- When should they stop and call a human?
- And, importantly, is the AI agent that was set up six months ago still doing something useful, or is it now producing a weekly report nobody reads?
On a grander scale, the AI Department may become responsible for something organisations have traditionally only done with people: workforce design.
Except this workforce will be part human, part machine and considerably less interested in the office Christmas party.
The AI Help Desk might sound a little different
The questions may also change.
Instead of: “Have you tried turning it off and on again?” we may hear: “Have you checked whether a human should be doing that?”
Employees might contact the AI team because their agent has gone rogue, their automated workflow has become more complicated than the process it replaced, or nobody can remember why an AI tool was introduced in the first place.
And perhaps that will become the department’s most important functions: stopping organisations from using AI simply wrong.
Part governance, part training, part reality check.
A successful AI Department could bring together several functions that are currently scattered across organisations.
It might oversee approved AI tools and agents, establish governance and human checkpoints, identify useful applications, train employees, review existing AI practices and help teams redesign work. Most importantly, it could provide a central place for organisations to learn from their own AI use.
- What is working?
- What is saving time?
- What is creating new problems?
- What should be scaled?
- What should be quietly switched off and never mentioned again?
Because AI adoption will not be a one-time technology rollout. It will require ongoing experimentation, evaluation and adjustment.
The real job of the AI Department
The biggest mistake would be to create an AI Department and make AI entirely its responsibility.
Just as cybersecurity is not solely the responsibility of IT, effective AI use will ultimately require capability across the entire organisation.
The AI Department should never become the place where everyone sends anything vaguely AI-related to avoid thinking about it themselves.
Its role should be to help the organisation develop the systems, skills and judgement required to use AI well.
Future organisational charts will include a Chief AI Officer, AI governance specialists, agent managers, AI trainers and workflow designers.
Or perhaps AI capability will eventually become so embedded in every department that a separate AI function is no longer necessary.
But during the transition, organisations may need people whose job is to ask AI questions everyone else is too busy to ask.
Should we automate this?
Who is accountable?
Where does the human need to remain involved?
And is this actually making work better?
Because the future AI Department may not exist simply to help organisations use more AI.
Its most important job may be helping them use the right AI, in the right places, for the right reasons.