September 17, 2026

AI Isn’t Taking Jobs. Bad Planning Is

Last week, I spoke to a hiring leader who said something I’m hearing more and more: ‘We’re not sure whether to hire two analysts, one AI engineer, or just wait six months and see what the tools can do.’

That right there is the issue.

Not killer robots. Not machine overlords. Not a dramatic end to civilisation narrated by someone with a podcast and too much free time.

The real story is simpler and messier. People, especially younger workers, are looking at AI and wondering whether the first few rungs of the career ladder are disappearing before they have even got a foot on it.

And if you’re running a business, hiring a team, or trying to build capability in data, tech, infrastructure or cyber, you need to take that seriously.

Because whether AI is actually ‘taking jobs’ is almost beside the point. If your workforce believes the rules have changed overnight, behaviour changes overnight too.

That means hesitation. Lower confidence. More noise in the talent market. More candidates second-guessing career choices. More employers delaying decisions while pretending it’s all part of some grand strategy.

It usually isn’t.

TechRadar’s piece gets at this tension nicely: the public conversation is obsessed with extinction-level AI risk, while many workers are more worried about something far more immediate, like whether there’ll be a job worth having in three years’ time. Source: TechRadar, ‘Forget AI ending humanity; what people are really worried about is AI taking their jobs — even if that's not exactly what's happening’.

That concern is not irrational. But it does need decoding properly.

What people are actually scared of

Most people are not worried AI will replace every job tomorrow morning before their first coffee.

They’re worried about three subtler things.

The entry-level squeeze.

If AI can draft, summarise, analyse, code, document and automate chunks of junior work, then people naturally ask: how do I get started in the first place?

That is a fair question.

A lot of careers in BI, data, software, cyber and operations are built on doing the unglamorous groundwork first. You learn by repetition. You spot patterns. You build judgment. If businesses automate too much of that layer without redesigning development paths, they create a nasty paradox: fewer junior roles now, fewer experienced leaders later.

The moving target problem.

Young professionals are being told to specialise, stay adaptable, build technical depth, learn AI, think strategically, communicate well, and somehow not panic. It’s like being handed a satnav that keeps rerouting every 90 seconds.

People do not mind learning. They mind feeling that the finish line moves every time they get close.

The trust gap.

Many employers are saying, ‘AI won’t replace people, it will augment them.’ Fine. Lovely phrase. Very conference-friendly.

But if that same business has a hiring freeze, is trimming apprenticeships, and expects one overstretched team to do more with automation, employees aren’t hearing ‘augmentation’. They’re hearing ‘good luck’.

AI is not replacing people. It is replacing poorly designed work

Here’s my take: AI is not mainly a job-destruction story. It’s a job-redesign story.

That is less cinematic, but a lot more useful.

The businesses winning with AI are not simply deleting headcount and calling it innovation. They are getting sharper about what humans should do, what machines should do, and what needs rethinking from scratch.

That distinction matters.

If a data analyst spends half the week pulling reports manually, cleaning spreadsheets and formatting slides no one reads, then yes, AI and automation should absolutely eat that work alive. Frankly, it should have happened years ago.

But the value of the analyst was never in the spreadsheet janitorial work. It was in asking better questions, spotting commercial patterns, influencing decisions and knowing when the data smells off.

Same story in infrastructure, cyber, engineering, and support functions.

AI often removes the repetitive layer. It does not remove the need for judgment. In many cases, it raises the premium on judgment.

That should be good news.

Unless, of course, your organisation has no plan for helping people make that leap.

The real risk is lazy leadership

This is the bit we should talk about more.

AI is exposing weak workforce planning the way low tide exposes bad beach architecture. Suddenly everyone can see what was flimsy all along.

Too many businesses still hire reactively, define roles vaguely, and treat capability-building like a nice-to-have until something breaks.

Then AI arrives, and instead of getting strategic, they do one of three things:

• Freeze hiring and hope the tools sort it out

• Buy AI platforms because everyone else is doing it

• Rewrite job specs with impossible wish lists and call it transformation

That is not a talent strategy. That is a mild panic in PowerPoint form.

If you’re serious about AI, you cannot dodge the people question.

You need to know:

• Which tasks are genuinely automatable

• Which roles need redesigning

• Which capabilities you must hire externally

• Which talent you can upskill internally

• How junior talent will still enter and grow in your business

Miss that last one and you’re basically eating your seed corn while congratulating yourself on efficiency.

Why this matters for hiring right now

From a recruitment perspective, AI anxiety is already changing the market.

Candidates are asking sharper questions. Good ones want to know whether a role has real longevity, whether the company is investing in people, and whether AI is being used with intent or just sprinkled over chaos like oregano on a bad pizza.

And employers? Some are clearer than ever. Others are paralysed.

The best hiring leaders I speak to are doing a few things differently.

They hire for adaptability, not just tool knowledge.

Today’s hot AI stack can become tomorrow’s forgotten tab. What lasts is structured thinking, commercial judgment, communication, curiosity and the ability to learn quickly.

They separate tasks from roles.

Just because AI can do 25 percent of a role does not mean the role disappears. It means the role changes. That sounds obvious, yet plenty of firms still think in a strangely binary way.

They are honest with candidates.

If AI will change how the team works, say that. Strong candidates are not scared by change. They are scared by fluff.

They protect the junior pipeline.

This one is huge. If your business stops hiring juniors because AI can cover some early tasks, you may look efficient this quarter and clueless in three years.

There is no shortcut around building future capability.

A simple framework for leaders

If you’re trying to make sensible hiring decisions in the middle of AI noise, use this four-part lens:

Automate

Identify repetitive, rules-based, low-judgment tasks that tools can handle faster or better.

Augment

Pinpoint roles where AI can improve output, speed or quality, but where human judgment remains central.

Apprentice

Protect the work and learning experiences junior talent need in order to become tomorrow’s senior talent.

Align

Make sure hiring, L&D, team design and technology strategy are connected. If those are operating in silos, you are not doing transformation. You are doing expensive drift.

Run every key role through those four lenses and you’ll get a much better conversation than ‘will AI replace this job?’

Questions worth asking internally

If you’re a founder, CTO, COO, CIO or people leader, take these into your next planning session:

• Which parts of our current roles are actual value creation, and which are admin theatre?

• Are we reducing low-level tasks without reducing learning opportunities?

• Have we redesigned roles, or are we just piling AI tools onto old workflows?

• What capability do we need 24 months from now that we are not building today?

• Can our junior talent still see a path forward here?

• Are we hiring for resilience and judgment, or for a shopping list of buzzwords?

Those answers will tell you more about your future workforce than any dramatic headline about AI ever will.

My blunt advice to candidates and employers

For candidates, especially younger ones: do not build your identity around tasks that can be templated. Build it around judgment, communication, domain understanding and adaptability. Learn the tools, yes. But do not become the tool.

For employers: stop using uncertainty as an excuse to sit on your hands. If AI is changing your operating model, say what that means for hiring, development and progression. Ambiguity is expensive. Good people can smell it a mile off.

And if your entire workforce strategy amounts to ‘let’s wait and see’, I can tell you how that ends. Usually with missed hires, patchy capability, and a desperate search six months later for someone who can magically sort out the mess.

That person does not usually arrive cheap, by the way.

Conclusion

The biggest misconception in the AI jobs debate is that this is about machines versus humans.

It isn’t.

It’s about whether leaders are prepared to redesign work, invest in people and make sharper hiring decisions while the ground is shifting.

AI is not the villain of the piece. Bad planning is.

The firms that win won’t be the ones shouting loudest about transformation. They’ll be the ones building teams that know what to automate, what to protect, and what human talent is actually for.

That takes thought. It takes guts. And yes, occasionally it takes calling in someone who can spot the difference between a future-proof hire and a very expensive guessing game.

Funny that.

Back to news