AI is already changing the workforce. Just maybe not in the way everyone expected.
We’re not seeing entire teams disappear overnight. We’re seeing fewer people being asked to do more.
Product managers are prototyping. Engineers are moving from brief to production. Marketing and GTM teams are building automated workflows that would once have passed through several people.
Job descriptions are stretching too. Businesses want people who can work across functions, make decisions with less direction and take something from idea to outcome without five handovers along the way.
On paper, it sounds great. Smaller team. Faster decisions. More output.
But at what point does a lean team just become an understaffed one?
One person is being asked to own a lot more
Between May and July 2026, Onset’s FY26/27 Hiring & Salary Insights Survey gathered views from professionals across technology, product, data, AI, GTM and leadership. The findings offer a directional snapshot, supported by the conversations we’re having across the market.
One theme kept coming up: AI is reducing the number of people businesses think they need, rather than creating new roles.
Respondents described businesses trying to maintain, or even increase, output with fewer people. One put it pretty bluntly: a business can operate “permanently understaffed” if the people left are strong enough.
That’s where the efficiency story gets more complicated.
There’s nothing wrong with hiring people who can have a bigger impact. The problem starts when “high impact” quietly becomes shorthand for one person doing what used to be three jobs.
That creates three risks worth paying attention to.
1. When too much depends on too few people
Imagine one person knows the customer context, the technical logic, the AI workflows and all the weird little things nobody thought to document.
They might be incredibly productive. They’re also a pretty big single point of failure.
If they leave or burn out, you’re not simply replacing one employee. You’re trying to piece together an entire way of working that lived mostly in their head.
Retention could become a problem too.
In our survey, 26% of respondents rated their salary satisfaction at four out of ten or below. Another 32% gave it a five or six.
That doesn’t mean they’re all job hunting. But if someone’s role has quietly doubled while their salary, title or recognition hasn’t moved with it, another offer can start looking pretty interesting.
If your new team dynamic only works while a handful of your best people are running at 100%, it probably isn’t as lean as it looks.
It’s fragile.
2. AI use is moving faster than the rules
Among the regular AI users who responded, 51% said they were already using tools their employer didn’t provide.
When we asked which skills were becoming more valuable in their field, AI fluency appeared in more than 80% of responses across every function represented.
The survey doesn’t tell us every business is in the same position. It does suggest employees may be adopting AI faster than their organisations are setting clear rules around it.
People weren’t just talking about prompting. What mattered was knowing how to apply AI: building workflows, checking outputs and understanding the risks.
Because the question isn’t really whether employees should be using AI.
They already are.
The more useful questions are: does the business know what tools people are using? What information is going into them? Who is checking the output? And if an automated workflow gets something wrong, who owns that mistake?
And as AI moves beyond simple prompts and starts researching, communicating and acting across several steps, those questions become a lot more important.
Being good at AI is useful.
Being good at AI without knowing how to check it can create a whole new category of problems.
3. We could be creating efficiency now and creating a talent problem later
The current market has made it easier for employers to ask for more experience at almost every level. You can understand why.
If an experienced person is available, why take a chance on someone who still needs developing?
The problem is what happens if everyone thinks that way.
Junior work does two things. It gets work done, but it also gives people the experience they need to become good at their jobs.
AI can take away some of that early work. It can’t magically create experienced people five years from now.
If businesses stop hiring and developing junior talent, the need for senior people doesn’t disappear. The path that creates those senior people does.
The businesses that keep investing in emerging talent could end up with something increasingly hard to find: people who actually understand the fundamentals of their profession and know when, where and how to use AI on top of them.
Don’t hire the unicorn. Build the right environment.
The hiring briefs are getting ambitious.
AI skills. Commercial judgment. Technical knowledge. Great communicator. Strategic thinker. Hands-on executor. Cross-functional. Ten years’ experience.
And, naturally, within budget.
There’s a point where a “high-impact hire” becomes one person compensating for everything missing around them.
AI will let smaller teams do more. Roles will get broader. Some work that once needed several people won’t anymore.
But fewer people shouldn’t automatically be the goal.
If the trade-off is exhausted high performers, important knowledge held by one person, AI tools nobody is tracking and fewer chances for people to learn their craft, the cost hasn’t disappeared.
It’s just moved somewhere less obvious.
That’s capability debt.
And eventually, someone has to pay it back.
Conclusion
Across Onset and The Omnia Collective, we help businesses work out where AI can genuinely improve the work, what should still sit with people and which skills their teams will need next. We see both sides: the people businesses are hiring, and the AI, data and technology changing how teams work. That helps leaders decide where AI adds value, where the risks sit and how to build a leaner team without becoming too dependent on a few people. If your business is working through this shift, that’s exactly the kind of conversation we’re having with teams right now.
For more insights into the forces shaping hiring, salaries and teams, download our FY26/27 Hiring & Salary Guide here.
