AI Can Make HR Faster. But Does It Make HR More Effective?

I use AI almost every day now, and there is no question that it has changed the way I work. I use it when I research a topic, develop an idea, structure information, analyse something, prepare a first draft or simply need another perspective on a problem I am working through. It saves me time. A lot of it, sometimes.

There are things that used to take me an hour that I can now do much faster. There are also things I can create or explore today that I probably wouldn't have attempted before because they would have taken too much time. I don't want to underestimate that benefit. AI is an incredibly useful working tool, and I think we are still discovering how much it can help us. But the more I use it, the more I notice something else. Being faster and being effective are not necessarily the same thing. That thought has changed the way I look at AI in HR.

Sometimes we are simply doing the same work faster

Imagine an HR team preparing a monthly report. It takes four hours to collect the information, organise it, write the commentary and prepare it for management. With AI, perhaps that same report can be prepared in one hour. That is a real improvement. Three hours have been saved, and I would absolutely take those three hours. But there is another question I would want to ask: what happens with the report?

If managers use it to understand what is happening in the organisation and make better decisions, then making it faster is valuable. But if hardly anybody reads it and nothing changes because of it, we have simply become much more efficient at producing something that creates very little value.

The same thing can happen with HR processes. A process might have seven steps, several approvals and information moving between different people. AI and automation can help us move through those seven steps much faster. But perhaps, if we looked at the process properly, we would discover that we only need four of them.

I have seen this happen with processes over the years. They rarely become complicated overnight. A new control is added because something happened once. Another approval is introduced. Somebody creates an additional spreadsheet. A report is requested and then continues to be produced long after anyone remembers why it was originally needed. Each addition may have made perfect sense at the time, but eventually the whole process becomes heavier than it needs to be.

Now we have AI, and there is a temptation to put this incredibly powerful technology on top of everything we already do. That can certainly make the existing work faster. But it doesn't necessarily make the work better.

Effectiveness starts before AI

This is actually not an AI principle for me. It is an HR principle. When I work on an HR process, policy, report, tool or initiative, I want to understand why we are doing it in the first place. What problem are we trying to solve? What outcome do we want? Who are we creating this for, and what should be different when we are finished?

I think these questions become even more important now that AI makes it so easy to create things. We can produce another policy, another presentation, another manager guide, another analysis or another communication incredibly quickly. But the fact that something has become easier to produce doesn't automatically mean we need more of it.

If I were reviewing an HR process today, I wouldn't begin by searching for places where AI could be inserted. I would first look at the process itself. I would try to understand where people are waiting, where information is duplicated, which steps genuinely protect the organisation or improve the outcome, and which ones have simply accumulated over time. I would look at what employees and managers actually need from the process and what HR is trying to achieve. Then I would simplify it. Only after that would I start deciding what people should do, what could be automated and where AI could genuinely improve the result.

Interestingly, newer research is increasingly pointing in the same direction. CIPD's 2026 research on how AI is changing HR capabilities highlights job design, workforce planning and organisational change as increasingly important capabilities as AI changes work. One of the challenges emerging from the research is that organisations can find it easier to introduce AI into existing work than to redesign jobs and work around the possibilities AI creates.

Gartner's 2026 work on redesigning HR processes for AI makes a related argument: increasingly sophisticated AI will not deliver its full potential if organisations simply add it to existing HR processes without reconsidering the processes themselves. That makes a lot of sense, technology can accelerate a process, but understanding the work is what allows us to improve it.

AI becomes much more powerful when we know what we're doing

None of this makes me want to use AI less. Quite the opposite. The more I understand what I want to achieve, the more useful AI becomes. When I know the subject, understand the context and have a clear idea of the outcome I want, I can give AI better direction. I can recognise when the answer is too generic, when something important is missing, when the logic isn't right or when the result simply doesn't fit the organisation.

This is why I think the conversation about AI literacy needs to go much further than learning how to write good prompts. For HR professionals, AI literacy also means understanding what the technology can and cannot do, knowing how to evaluate what it produces, recognising its risks and limitations, and understanding when human judgement needs to take over. [Related Insight: What AI Literacy Actually Means for HR →]

The difference becomes especially important when we move away from creating documents and start using AI around decisions about people.

Because HR decisions affect real lives

This is where I think HR needs to be particularly thoughtful. The information we work with isn't abstract. Behind the data are people whose lives can be affected by the decisions an organisation makes.

Recruitment influences whether somebody gets an opportunity. Pay decisions affect people's income. Performance assessments can influence careers. Decisions about promotion, development, restructuring or termination can have consequences far beyond an HR system. AI can be extremely valuable around these activities. It can help organise information, identify patterns, prepare analysis and support people making decisions. But supporting a decision and making the decision are very different things.

This is also reflected in regulation. Under the EU AI Act, certain AI systems used in employment and worker management — including some applications connected with recruitment, selection, promotion, termination, task allocation and performance monitoring — are classified as high-risk because of the potential impact on people's rights, careers and livelihoods. European Commission — AI Act regulatory framework

For me, governance isn't something we should introduce after we have already implemented AI. It needs to develop alongside AI use. We need to understand what information we are giving these systems, how outputs are being produced and used, who checks them, who can challenge them and, ultimately, who remains responsible for the decision. Because a system can process information. It doesn't carry responsibility for what happens to the person afterwards. We do.

And then there is the time AI gives us back

This may actually be the part I find most interesting. If AI saves me five hours this week, what do I do with those five hours? It would be very easy to simply fill them with more work. Another document, another analysis, another meeting, another project. Before long, I would be just as busy as I was before, only producing more things. But there is another possibility.

Perhaps some of that time could be used to look again at an HR process we haven't questioned for years. To understand why managers are struggling with something instead of sending them another guide. To look properly at employee data rather than simply producing the dashboard. To improve onboarding instead of administering it. To develop managers, learn something new, talk to employees or simply have enough space to think properly about a problem before jumping into the solution.

Anyone who has worked in HR knows how easily important work gets pushed aside by urgent work. There is always a vacancy, a contract, an employee issue, a deadline or a manager who needs something now. AI has the potential to create some space between those things. But that only becomes valuable if we make conscious choices about what we do with the space.

The ILO's 2026 review of evidence on generative AI, productivity and work organisation is particularly interesting in this context. It finds genuine productivity benefits from generative AI, but also considerable differences depending on the task, worker and context. Importantly, reported time savings have not automatically translated into broader improvements in measured outcomes.

That reinforces something I have increasingly felt through my own use of AI: time saved is valuable, but it is not the final outcome. What we do with that time matters.

Perhaps this is where HR has a bigger role to play

The more organisations use AI, the more I think HR's role in this transition could become much broader than helping HR teams use new tools. AI is changing work itself. Jobs will change as some tasks become automated and others become more important. Employees will need different skills. Managers will need to learn how to lead people whose work increasingly involves AI. Organisations will need to decide where human judgement remains essential, how performance should be measured and how productivity gains should be shared without simply turning them into an expectation that everybody should do more. Those are not purely technology questions.

They are questions about work, people, capability, leadership and organisational design. In other words, they are very much HR questions. And perhaps that is one of the most interesting opportunities AI creates for our profession. Not simply becoming better users of AI, but helping organisations make better decisions about how people and AI should work together.

The question to start with

When I look at a work today, I try not to begin with Where can I use AI? I want to understand what I am trying to achieve first. Why are we doing this? What outcome do we need? Who are we doing it for? What actually needs to happen to achieve that outcome?

Once that is clear, I can look at the work itself. I can remove what doesn't add value, simplify what has become unnecessarily complicated, decide where human expertise and judgement matter, and then use technology, automation and AI wherever they can genuinely make the work better or faster.

I absolutely want AI to save time. I want it to remove repetitive work, help me think, make research easier, accelerate analysis and reduce the hours spent doing things technology can do perfectly well. But I don't want us to confuse doing more with becoming more effective.

AI can do amazing things when it is used by people who understand what they are trying to achieve, know how to work with the technology and remain responsible for the outcome.

Not AI instead of good HR, and not good HR instead of AI. Good HR thinking first — and AI used intelligently to make it even better.