%20-%20Thumbnail%20(1).png)
There is a version of AI adoption that looks great on a slide deck.
Enthusiastic individuals experimenting with new tools. Leaders discussing transformation at conferences. Pilots running in pockets of the business.
And then there is what is happening inside most HR teams day to day.
Onboarding still depends on someone remembering to send the right email. Approvals sit in inboxes. Processes that should take hours take days, because the steps between systems are still managed by whoever is around to manage them.
We surveyed 200 UK HR leaders to understand the gap between the two. Here is what the data actually shows.
71% of HR professionals use AI daily or weekly.
Only 17% work in organisations where automation is extensively built into most HR processes.
That gap is not a surprise to anyone working in HR right now. But it is worth naming clearly, because it points to something specific.
Using AI to draft a job description or summarise a document is useful. It makes you faster. But it does not change what happens when a new starter joins, when a manager needs to approve a request or when HR needs to coordinate across payroll, IT and facilities at the same time.
Personal AI use and workflow automation are two different things. Most organisations have made progress on the first. Far fewer have tackled the second.
97% of HR teams experience errors or delays from manual processes at least once a month.
For nearly half, those errors happen weekly or daily.
That is not an edge case. It is the operating reality for most HR teams, quietly running in the background of every working week.
Some of the cost is obvious. Time spent chasing approvals. Mistakes that need correcting. The same policy question answered three different ways by three different people.
Some of it is less visible.
HR professionals spending 20 to 30 hours a week on admin are significantly more likely to report near-constant stress than those spending 10 to 20 hours. Over 52% of those with heavy admin loads say their workload contributes to stress frequently.
This is not just a productivity conversation. It is a wellbeing conversation, a retention conversation and a capacity conversation. When HR teams are buried in manual work, the strategic stuff gets squeezed out.
When a process breaks down, the instinct is usually to add something new.
Our data suggests that approach is backfiring.
Organisations using five or more HR tools report weekly errors at a rate of 45.2%. Those using five to six tools see that spike to 52.2%. Organisations using three to four tools sit at 22.2%.
More tools without proper connections between them means more manual work filling the gaps. Someone has to move information from one system to another. Someone has to reconcile what one platform says with what another shows. That someone is usually in HR.
Integration with your existing tools is not a nice-to-have. It is the thing that determines whether automation actually works or just creates a new layer of admin.
When we asked HR leaders where their time goes, five areas came up consistently.
Recruitment and onboarding accounts for the largest share at 26.5%, followed by compliance and documentation at 21.5%, training and development at 19.5%, performance management at 16% and employee requests and case management at 9%.
What these have in common is not just that they are time-consuming. It is that all five rely on coordination across multiple systems and multiple people, with no single process holding them together.
That dependency is where the delays come from. It is where errors happen. And it is where a large chunk of the working week quietly disappears.
Here is the telling stat: if admin work in HR were reduced by 25%, only 3% of leaders say their team would continue focusing mainly on administration. 39% would shift to strategic workforce planning. 29.5% would prioritise improving employee experience.
The work HR teams want to be doing exists. It is just buried under the work that has not been properly automated yet.
77% of HR leaders plan to prioritise automation in the next 12 months.
22.1% have fully implemented it.
Compliance concerns come up most often as a barrier, cited by 35.6% of organisations. Budget constraints follow at 25.3%, with disconnected tools at 20.7%.
But when we looked at organisations actively piloting automation, a different challenge kept appearing: data quality and access. 32.2% said this was their most immediate hurdle. Before you can automate a process reliably, the right information needs to flow between the right systems. For many teams, that foundation is not quite there yet.
There is also a structural problem that does not show up cleanly in surveys.
Most HR teams cannot build or change workflows themselves. When a process needs updating, it joins a queue. When a new tool needs connecting, it becomes an IT project. The people who understand the process best are rarely the people with access to improve it. That disconnect slows everything down.
The 17% who have genuinely embedded automation are not necessarily the ones with the biggest budgets.
They tend to do three things well.
They automate the handoffs between steps, not just individual tasks within them. An onboarding process where someone still manually triggers IT access is not really automated. One where access is provisioned automatically when a contract is signed is a different thing entirely.
They give HR teams the ability to build and update workflows without depending on technical support. When every change requires a ticket, momentum dies.
And they treat compliance as something the process handles by design, not something checked at the end. Approval chains are defined upfront. Audit trails happen automatically. The workflow itself becomes the evidence.
AI works well here as a co-pilot, helping HR teams describe what they need, build the logic and refine it over time, rather than replacing the people who understand how the process should actually work.
The organisations making the most progress tend not to have the most ambitious plans.
They picked a specific process, understood how it works today and automated it properly before moving on to the next one.
Onboarding is the most common starting point. It touches the most systems, involves the most manual steps and has the most direct impact on how new employees experience the organisation in their first few weeks.
But the right starting point is usually wherever the friction is highest.
Somewhere in most HR teams there is a process that generates too many emails, depends on a spreadsheet nobody fully trusts or relies on one person knowing what comes next. That is usually where the most practical progress begins.
This article draws on findings from our 2026 industry report, The State of HR Automation: Bridging the Gap Between Intent and Impact, based on a survey of 200 UK HR leaders.
Read the full report
