Just 31% of UK organisations currently use AI in recruitment, up from 16% two years ago, according to CIPD’s own most recent data. That’s real growth, but it’s also a long way from the wholesale HR automation that headlines keep predicting. The honest picture sits somewhere between “nothing’s changing” and “HR is being replaced,” and most coverage of this question settles for one extreme or the other rather than doing the work to find out which.
This piece applies the same exposure framework already used throughout this series, specifically to HR: which parts of the job are genuinely automatable, what’s actually being automated right now, and why the compliance and judgement side of HR, the parts covered throughout this entire series, is if anything becoming more valuable, not less.
Explore: AI and the Future of Work: What’s Actually Happening to UK Jobs, for the broader picture. Explore also: Which Jobs and Skills Are Actually Changing and Career Guide: Future-Proofing Your Role in the AI Era, for the full framework this piece applies specifically to HR.
Where HR Actually Sits on the Exposure Scale
The government’s own exposure framework, already covered in detail elsewhere in this series, splits jobs by task rather than title:
- Boost-leaning tasks: more likely to be enhanced by AI than replaced
- Replace-leaning tasks: more likely to be performed by AI than a person
- Most roles are a genuine mix of both
HR maps onto this cleanly once you break it down by task rather than treating “HR” as one undifferentiated job. Recruitment screening, payroll administration, and routine policy queries lean toward “replace”, they’re repetitive, rule-based, and produce a predictable output. Employee relations casework, disciplinary judgement calls, and strategic workforce planning lean toward “boost”, they require judgement about ambiguous situations, and the consequences of getting them wrong are serious enough that a human needs to be making the actual call.
The government’s own caution applies here as much as anywhere else: exposure isn’t destiny. A recruitment coordinator whose role is mostly screening and scheduling sits in a genuinely different position than an HR business partner whose role is mostly employee relations and strategic advice, even though both carry the job title “HR.”
What’s Already Being Automated in HR (And What Isn’t)
The real, current data is more measured than either the hype or the anxiety suggests.
CIPD’s most recent Resourcing and Talent Planning Report found 31% of UK organisations now use AI or machine learning in recruitment, up from 16% in 2022. Where it’s being used, it’s concentrated in specific, fairly narrow tasks: writing job adverts, screening CVs against defined criteria, and responding to candidate queries through chatbots. Among organisations that have adopted it, 66% report improved hiring efficiency.
The honest counterbalance matters just as much. The same research found roughly 80% of SMEs, public sector organisations, and non-profits use no AI at all in their recruitment process. Adoption is real and growing, but it’s nowhere near universal, and it’s concentrated in larger, better-resourced organisations rather than spread evenly across the sector.
What remains stubbornly human is everything this series has spent the most time on: employee relations cases, disciplinary judgement calls, the kind of nuanced decisions covered throughout the discussion of AI-related grievances and discrimination claims elsewhere in this series. No credible data anywhere suggests AI is meaningfully automating this side of HR, and the legal framework covered throughout this series actively requires a human to be making these calls, not a system.
The Jobs Genuinely at Risk vs the Ones Being Redefined
The honest answer to “is AI coming for HR jobs” depends entirely on which HR job you mean.
Narrow, administrative HR roles, the kind built almost entirely around the tasks covered above, are genuinely exposed. A role that’s mostly CV screening, scheduling, and routine correspondence looks structurally similar to any other administrative role with high AI exposure, and the same pressures affecting administrative roles across the wider economy apply here too.
But that’s a different claim from “HR is shrinking.” Growth in AI-adjacent work is being driven less by entirely new jobs appearing and more by existing roles absorbing AI responsibility, already covered in detail elsewhere in this series. HR is a clear example of exactly this pattern. The administrative layer of the job is genuinely being automated, but that doesn’t eliminate the role, it shifts what the role is actually expected to do, less time on tasks a system can now handle, more time on the judgement-heavy work that was always the harder, more valuable part of the job anyway.
The practical distinction worth holding onto: a specific task within HR being automated is not the same as the HR function disappearing. It’s closer to the opposite, automating the administrative layer is what makes room for the strategic and compliance-heavy work that’s actually growing.
Why the Compliance and Judgement Side of HR Is Growing, Not Shrinking
This is where the case for HR’s growing importance moves from framework to evidence, because this series has already documented it in detail.
The Employment Rights Act 2025 alone has added genuine, ongoing complexity to HR’s role:
None of this is administrative work a system can absorb. All of it requires a human who understands both the legal framework and the specific situation in front of them.
AI itself has added a second layer of complexity rather than removing it. This series has covered:
Every one of these is a genuinely new area of judgement that didn’t exist in this form five years ago, and every one of them sits squarely inside HR’s remit, not outside it.
The pattern is consistent: as routine HR tasks become more automatable, the judgement, compliance, and legal literacy required to manage everything AI itself is introducing has grown right alongside it. HR isn’t being replaced by AI. It’s being asked to understand AI well enough to manage the risk it creates, on top of everything it already had to manage.
What This Means for HR Professionals
None of this points toward a single verdict on whether “AI is coming for HR jobs.” It points toward a more useful question: which parts of your own role are genuinely exposed, and which parts are becoming more valuable precisely because of what AI is introducing.
- Audit your own tasks honestly, the same exercise covered in detail elsewhere in this series. Administrative, repetitive tasks are where exposure is real. Judgement-heavy work is where it isn’t.
- Build genuine literacy in the areas AI itself is creating demand for: governance, data literacy, and the specific legal frameworks now governing AI-influenced decisions.
- Don’t treat automation of administrative tasks as a threat to resist. It’s what makes room for the strategic and compliance work that’s actually growing.
- Treat formal qualification as the credible way to demonstrate this capability, not just to build it privately.
How Avado Can Help
The judgement, compliance literacy, and legal understanding this piece points to, exactly what makes HR harder to automate, is what a genuine HR and L&D qualification is built to develop. Avado’s CIPD qualifications, from Level 3 through Level 7, build exactly this capability, backed by free 1:1 career mentorship and Avado’s CPD-accredited Career Accelerator programme.
Explore Avado’s CIPD courses and build the judgement that keeps HR indispensable, not automatable!
