Nine in ten CHROs say AI is their most pressing concern. Fewer than four in ten organisations are actually using it in HR. That gap is where the real work begins.
I've spent a good chunk of this month with the data on AI in HR. But I've also been talking to lots of HR leaders at very different stages of their AI maturity… some just getting started, others a long way down the road. Those conversations have shaped my thinking as much as any report, and between the two I keep coming back to one gap.
In the CHRO Association's 2026 survey, run with the University of South Carolina's Darla Moore School of Business, 91% of the roughly 150 CHROs asked put AI and the digitisation of the workplace among their five most immediate concerns. It ranked first, and nothing else came close.
Yet SHRM's State of AI in HR 2026 report, drawing on 1,908 HR professionals, finds that just 39% of organisations are using AI in their HR function today. They're different surveys, asking different people slightly different questions, so I won't pretend it's a like-for-like comparison. But the direction of travel is hard to miss… the priority we talk about and the practice we permit are still some way apart.
So, my verdict in two sentences. HR has moved. It hasn't transformed.
The movement is real, and I think it deserves credit. Employee-facing assistants now answer routine questions at two in the morning. Recruitment tools screen and shortlist at a pace no team of coordinators could match. Automation is stripping hours of admin friction out of HR operations, and early copilots are helping managers prepare for conversations they used to walk into cold. These things matter. They save time, and time is the scarcest thing any HR function has.
But look at where they sit. Almost every one of these tools is layered on top of the existing model. The processes are the same processes… they just run faster. SHRM's data shows adoption clustering in recruiting (27% of organisations), HR technology (21%) and learning and development (17%). In my experience, it's barely registering in the harder terrain of business partnering, reward or inclusion. Around six in ten organisations aren't using AI in HR at all. We're experimenting, often enthusiastically. We're not yet reimagining how HR work gets done.
The distinction matters because transformation has a shape. In three years, I think a genuinely AI-enabled HR function will feel structurally different, not just quicker. Transactional work will be handled end to end by intelligent systems. Insight will flow continuously, rather than arriving in periodic reports that describe a workforce that's already moved on. And HR people will spend less time producing, policing and processing, and more time interpreting, advising and stepping in where judgement, ethics and humanity matter most.
Getting there means letting go of some familiar comforts: control, manual assurance, and the quiet belief that HR proves its value through effort rather than impact. I understand those attachments – I've held a few of them myself. But they're also the brakes.
The CHRO data hints at the same thing. Nearly half (47%) say they haven't yet set clear productivity measures for AI. If we can't show what it's delivering, we'll struggle to make the case for the next round of investment, let alone a redesign of the function.
So far the visible benefits are efficiency gains, and most functions are stopping there. That's the smaller prize. The real opportunity is human. Used well, AI brings HR closer to what matters: employees getting timely, accurate answers; managers walking into difficult conversations properly prepared; HR teams spending their judgement on the moments that genuinely shape someone's experience of work. Efficiency buys the time. What we do with it is the transformation.
Now the turn, because the risks aren't hypothetical. The first is bias scaled rather than addressed. An unfair human decision harms one candidate at a time; an unfair algorithm can harm thousands before anyone notices. The second is tools deployed without enough clarity on ethics, accountability or long-term consequence… procured enthusiastically and governed thinly.
From a governance perspective, HR's job here isn't to slow progress. It's to make sure progress is deliberate. Being thoughtful isn't the same as being timid.
Which brings me to readiness, where the data is bluntest. In Workday's Global CHRO AI Indicator research, 42% of HR leaders said their team wasn't fully prepared for AI, the highest level of unpreparedness of any function surveyed. That's 2023 data, from 640 HR leaders within a wider sample of 2,355 senior executives, so things will have moved since. I'd be surprised if they've moved far enough.
And notice what kind of problem it is. The CHROs in this year's survey said the biggest barriers to scaling AI are organisational rather than technological: employee fear of job loss, budget constraints, and data, security and legal concerns. I get that it needs tech, of course. But I think this is mostly a gap of mindset, confidence and influence.
That reframes the task. HR has to see itself as a shaper of how AI is applied to people and work, not just a consumer of tools chosen by others. The function that understands work, behaviour and consequence better than any other should be setting the terms. If it doesn't, the transformation will happen anyway, on terms set by people who understand the technology far better than they understand people. That outcome is available. I wouldn't recommend it.
So, to fellow CPOs reading this… start with intent, not technology. Name the human and business problems you're solving before you name a vendor. Invest in literacy and confidence across your whole HR team, not just a handful of specialists, and make it a line item in next year's plan rather than an aspiration. Create space to experiment safely, learn quickly and course-correct early. This isn't a systems upgrade. It's a leadership moment for our profession.
That's a big part of why I built the HR AI Maturity Index. It's a free resource for the profession, not a commercial product: 28 questions across six pillars, from HR foundations and people data through to governance and compliance, organisational readiness and the role of the CPO. It's a diagnostic rather than a league table. It gives you an honest read of where your function is today and, more importantly, a personalised set of tangible next steps to turn intent into action. If it's useful, please do share it with your team and your peers.
FREE ASSESSMENT · 20–25 MINUTES
[Take the free assessment → www.hraimaturityindex.com](https://www.hraimaturityindex.com)
Nine in ten of us have said AI is the priority. Fewer than four in ten organisations are acting on it in HR. Which of those numbers will your function be explaining this time next year?
I might be missing something here, so please do shout. I'd love your perspectives.
Sources
CHRO Association with the University of South Carolina's Darla Moore School of Business, 2026 CHRO Survey: Key Findings (March 2026). Based on responses from an average of 150 CHROs at major corporations. www.chro.org/w/2026-chro-survey-key-findings-1
SHRM, The State of AI in HR 2026 (April 2026). Insights from 1,908 HR professionals. www.shrm.org/topics-tools/research/state-of-ai-hr-2026/full-report
Workday, Global CHRO AI Indicator Report: A Vision for Strategic Value (2023). Global survey of 2,355 senior executives conducted May–June 2023, including 640 HR respondents.