What the data says about HR in 2026

Every year we read the same trend list, and most of it does not survive February. It is worth looking at what actually changed level in 2026, with numbers.

AI left the pilot and entered the process

According to SHRM's research on the state of AI in HR, 87% of CHROs expect greater adoption of AI inside HR processes, up from 83% the previous year. Among organizations that have already implemented it, usage is no longer occasional: 26% use it weekly, 20% daily and 9% several times a day.

What interests me in that data is not the percentage. It is the change in the nature of the problem. When a tool is used several times a day, it stops being a project and becomes a process, and the whole change management repertoire starts to apply: who changes routine, who loses autonomy, who has to learn something new just to keep delivering the same thing.

Analytics stopped explaining the past

AIHR's reading of workforce analytics in 2026 describes the shift well: the function moved from explaining past outcomes to helping leaders anticipate risk and choose between alternatives.

In practice this changes the question a business partner receives. It used to be "what was turnover last quarter". Now it is "if we restructure this area, what is the risk of losing the people who hold the operation together". The second question is much harder and much more useful.

HR inside the technology strategy, not after it

This is the data point I use most in conversations with leadership: companies ahead in AI adoption are 2.5 times more likely to involve HR in helping employees identify which tasks are suited for automation. Involving HR from the start accelerates adoption and reduces resistance.

It makes sense up close. Resistance to a new tool is rarely about the tool. It is about what the tool signals: that the person will lose relevance, that the bar will rise without anyone explaining how, that the work they have always done no longer counts. None of that is solved with feature training.

What this changes in my work

Three things, concretely:

• Impact assessment now includes tasks, not only roles. With partial automation, job design does not change all at once: it changes task by task. The assessment has to follow at that level.
• Readiness became a continuous measure. In a traditional system project you measure readiness before go-live. With an AI tool, the usage curve keeps moving months later.
• The career conversation moved into the change plan. When part of the work is automated, the question "and what is left for me" shows up in week one. If the plan has no answer, the answer becomes rumour.

Where I disagree with the consensus

There is a narrative that HR needs to "become tech". I disagree. What HR needs is to keep doing well the thing that was always hard, which is understanding the impact of a decision on real people, and to have enough fluency to take part in the technical decision rather than watch it. Those are different things.

Sources: SHRM, The State of AI in HR 2026 and AIHR, Workforce Analytics Trends.