The 40 Jobs Most at Risk from AI...and What That Means for the Rest of Us
Sky News recently shared the results of an interesting Microsoft study examining which jobs are most, and least, at risk of being transformed by artificial intelligence (AI). The research analysed around 200,000 real conversations with Microsoft’s Copilot, assessing how much of each role’s workload could already be handled by AI tools.
The results are striking. Some jobs, it suggests, could see up to 90% of their tasks automated. Others, barely 2%. As someone working in recruitment, I can’t help but see this through the lens of skills, hiring, and future employability...and what these numbers might really mean for people in the job market.
The shifting landscape of work
The study found that jobs involving written analysis, communication, or digital creation are most vulnerable. Historians, coders and writers top the list, with more than 90% of their day-to-day work judged automatable. Roles such as customer service assistants (72%), financial advisers (69%) and product promoters (62%) also feature prominently among the “at risk” occupations.
In contrast, the jobs considered safest from AI are understandably those requiring hands-on skill, physical dexterity or human empathy. Roofers and cleaners have only around 2-3% of their tasks potentially affected, while nursing and surgical assistants sit at around 7%. These figures tell a clear story: the more human, manual or relational a role is, the harder it is to automate.
It’s also notable that, according to PwC data referenced in the article, job postings for the “more threatened” roles have grown four times slower than for those in the low-risk category between 2019 and 2024. At the same time, 78% of global businesses say they plan to increase AI spending this year, and around 40% expect to reduce headcount where automation can take over.
For all the optimism about productivity, the pace of change is sobering.
It’s not the job that’s at risk...it’s the task
One of the biggest misconceptions about automation is that AI comes for entire roles. In reality, it’s coming for the tasks that make up those roles. Two people with the same job title might face completely different futures depending on which parts of their work are repetitive or rules-based versus those that rely on judgment, creativity or trust.
That distinction is crucial for both employers and jobseekers. If we understand where the risk lies, we can redesign work around it. The winners won’t necessarily be in “safe” industries, they’ll be the people who adapt first, learning to partner with technology rather than compete against it.
🔌Shameless plug🔌 This is a topic I've discussed on 'A Quick Brew' podcast a number of times with my co-host James Sharples. Check it out for our thoughts on tech, careers, wellbeing, business and more! |
The polarisation problem
The article also hints at a trend that’s becoming hard to ignore: a growing divide between “augmented professionals” and “resilient hands-on” workers.
On one side, we have knowledge-based roles that will survive by integrating AI, the marketer using data-driven insights, the recruiter leveraging automation to screen but still building relationships. On the other, we have the trades, care and service roles that AI simply can’t replicate.
This could create a polarised labour market. One half heavily digital, the other profoundly human. And in between lies a disappearing middle, the routine administrative or procedural work that once provided stable employment for millions.
What this means for recruiters
For those of us in recruitment, this isn’t just a curiosity. It’s a call to action.
We need to start talking about automation resilience as a skill in itself. When evaluating candidates, it’s not enough to tick off qualifications or experience; we should be asking, how adaptable is this person? How do they respond to change?
Likewise, businesses need to think beyond short-term hiring needs and consider how their roles will evolve. Could a task-heavy role be redesigned to focus on oversight, creativity or client engagement instead? Could an existing employee be trained to work alongside AI rather than replaced by it?
And while much of the public debate focuses on white-collar automation, we shouldn’t overlook the enduring importance of practical, people-centred work. The very jobs many overlook, trades, healthcare support, logistics, maintenance, these could soon become the most stable careers in an increasingly digital economy.
My take
The Sky piece isn’t predicting an overnight apocalypse. It’s a snapshot of how far the technology has already reached. But it should make all of us - recruiters, employers, and jobseekers, pause and reflect.
AI will almost certainly reshape the job market, but that doesn’t mean the human element disappears. In fact, it may become more valuable. The qualities machines can’t replicate like empathy, judgment, ethics, persuasion, and trust will define the future of employability.
For me, the real question isn’t “Will my job survive?”...It’s “What can I do that AI can’t. And how can I build a career around that?”
If we can help people answer that, the recruitment industry won’t just survive the AI revolution. It may just help to lead it.