AI in People Analytics: Top 3 Questions Organizations Should Ask as AI Agents Join the Workforce

For years, organizations have understood their workforce through familiar questions: How many employees do we have? What roles do they perform? What skills do they possess? How productive are they? Where do we need more people? 

Artificial intelligence is beginning to complicate those questions. 

AI is no longer being used only to write an email, summarize a document, or answer a question. More advanced AI agents can carry out several steps of a task, use information from different sources, and complete parts of a workflow with human direction and oversight. 

Microsoft’s 2026 Work Trend Index describes this shift as AI taking on more of the execution while people increasingly direct the work, exercise judgment, and remain accountable for the outcome. Its research, based on 20,000 AI users across 10 countries, suggests that the challenge is moving beyond simply giving employees access to AI. Organizations now need to reconsider how work itself is designed. 

That creates an interesting question for HR. 

If part of the work is now being performed by AI, do our traditional ways of understanding the workforce still tell us enough? 

This is where AI in People Analytics takes on a new role. It is no longer only about analyzing employee information. It can also help organizations understand how people, skills, roles, and increasingly AI come together to deliver work. 

As this shift develops, here are three questions organizations should begin asking.

1. What Work Should People Do, and What Work Should AI Do?

When organizations talk about AI and the workforce, the discussion often jumps immediately to jobs. 

Will AI replace accountants? Recruiters? Analysts? Customer service representatives? 

But a job is actually made up of many different activities, and AI may affect each of those activities differently. 

Consider a recruiter. A recruiter may source candidates, review applications, prepare interview questions, communicate with hiring managers, interview candidates, evaluate fit, negotiate offers, and advise leaders on talent decisions. 

AI may be able to assist significantly with sourcing, summarizing applications, preparing information, or administrative follow-ups. However, evaluating a sensitive situation, building trust with a candidate, challenging a hiring manager’s assumptions, or making a difficult judgment may still require much stronger human involvement. 

This means organizations may need to look beyond jobs and begin looking more closely at work. 

Microsoft’s recent research describes advanced AI users as deliberately deciding which activities should be delegated to AI and which require greater human involvement. It also found that 53% of its most advanced AI users said they intentionally pause to decide whether a task should be done by a person or by AI, compared with 33% of other AI users. 

For People Analytics, this creates a new area of inquiry. 

Instead of asking only, “How many people do we need for this function?”, organizations may increasingly need to ask, “What work needs to be done, which parts can AI support, and where does human judgment create the most value?” 

That is a very different workforce planning conversation. 

It could also prevent organizations from making the mistake of introducing AI simply because a task can be automated, without considering whether it should be. 

2. What Skills Become More Important When AI Can Do More of the Execution?

AI does not only change how much work people can accomplish. It may also change which human capabilities matter most. 

Imagine that an analyst previously spent several hours gathering information, organizing data, preparing a first draft, and creating a presentation. AI may now help complete some of those activities much faster. That does not necessarily make the analyst less important. 

Instead, the analyst’s value may move toward different questions: 

• Can they identify the right business problem? 
• Can they tell whether the AI’s analysis actually makes sense? 
• Can they recognize missing information? 
• Can they challenge an incorrect conclusion? 
• Can they explain the result to a decision-maker? 
• Can they take responsibility for the recommendation? 

Microsoft’s 2026 research reflects this shift. Among the AI users it surveyed, quality control of AI outputs and critical thinking were the two most commonly identified human skills becoming more important as AI takes on more work. 

The broader skills landscape is also changing quickly. The World Economic Forum reports that employers expect nearly 40% of skills required on the job to change by 2030, while technology skills and human capabilities such as analytical thinking, leadership, resilience, and collaboration are all expected to remain important. 

For organizations, this creates an important People Analytics question: Are we developing employees for the work they perform today, or for the work they will need to perform alongside AI? 

A company may have enough employees on paper and still face a capability problem if those employees are not prepared to supervise AI, evaluate its outputs, redesign workflows, exercise judgment, or apply technology responsibly. 

This is why understanding the workforce increasingly requires more than tracking headcount or job titles. Organizations need better visibility into the skills behind the roles and how those skills are changing. 

3. Are We Still Measuring Workforce Capacity and Productivity the Right Way?

This may be the most significant question of all. For decades, organizations have relied heavily on measures such as headcount, hours worked, output per employee, and the number of people assigned to a function. 

Those measures are still useful, but what happens when AI contributes meaningfully to the work? 

Imagine two teams with ten employees each. The first team uses AI mainly for occasional drafting and research. The second has redesigned several workflows so that AI handles routine information gathering, prepares initial analyses, automates administrative work, and supports employees in completing more complex tasks. 

Both teams have the same headcount. But do they have the same capacity? Probably not. 

Microsoft reported in 2026 that 66% of surveyed AI users said AI allowed them to spend more time on higher-value work, while 58% said they were producing work they could not have produced a year earlier. 

This suggests that headcount alone may tell organizations less about workforce capacity than it once did. 

People Analytics may therefore need to examine a wider set of questions. 

• How much work is being augmented by AI? 
• Which activities have become faster? 
• Has the quality of the output improved? 
• Are employees actually spending the time saved on more valuable work? 
• Are workloads becoming healthier, or are employees simply expected to produce even more? 
• Are some teams benefiting from AI considerably more than others? 
• And most importantly, are these changes producing better business results? 

This distinction matters because AI adoption is not automatically the same as AI value. 

A company could have thousands of employees using AI every week and still see very little meaningful improvement if the underlying work has not changed. 

People Analytics can help organizations move the conversation from: 

“How many employees are using AI?” to: “What is changing because they are using it?” 

Why This Matters to Organizations Today

The introduction of AI into the workforce is not simply an IT project. 

It affects how organizations think about workforce planning, skills, productivity, organizational design, learning, recruitment, performance, and leadership. 

Microsoft’s 2025 Work Trend Index found that 45% of surveyed leaders were considering expanding team capacity through what it calls digital labor, while 47% identified upskilling existing employees as a priority. It also introduced the idea of a “human-agent ratio”: deciding where work should be automated, where people and AI should collaborate, and where stronger human involvement remains necessary. 

That concept is particularly relevant to HR. 

The workforce question is gradually becoming more complex than: How many people do we need? 

It may increasingly become: What outcomes do we need to deliver, what capabilities are required, what should people do, what can AI support, and how should the two work together? 

Organizations that cannot answer those questions may invest heavily in AI without truly understanding whether it is strengthening workforce capacity. Organizations that can answer them have an opportunity to make AI adoption much more intentional. 

How ASEAMETRICS Can Help

This is where People Analytics can help organizations bring the workforce conversation back to evidence. 

Through PeopleAnalytics, ASEAMETRICS can help organizations better understand the workforce they have today by connecting relevant people and business information, identifying capability and workforce patterns, establishing baselines, and determining where closer investigation may be needed. 

For example, an organization considering greater AI adoption may need to understand which roles are changing, where capability gaps exist, which employee populations require development, how workforce capacity is shifting, and whether changes in how work is performed are actually improving results. 

Those insights can then inform the decisions leaders make about workforce design, development, reskilling, redeployment, or other interventions. 

Through M&Ensight, organizations can manage what happens after the insight is identified by defining interventions, assigning accountable owners, setting milestones and measures, documenting expected benefits, and tracking whether the intervention is producing the intended organizational result. 

The cycle remains straightforward: 

Business Priority → Workforce Insight → Decision → Intervention → Accountability → Measurement → Business Result 

Organizations do not need to redesign the entire workforce immediately. 

They can begin with one business process, one employee population, or one workforce question. 

For example: 

Where is AI changing the way work is performed in our organization, and do our people have the capabilities to work effectively in that new environment? 

Answering one question like this can create a much clearer starting point for understanding what the future workforce actually needs. 

Because as AI becomes part of how work gets done, the most important People Analytics question may no longer be simply: 

“Do we have the right number of people?” 

It may be: 

“Do we have the right combination of people, skills, technology, and ways of working to deliver what the organization needs next?”