AI In Human Resource Management

AI Is Changing How People Work. Here’s How HR Can Redesign Jobs

Shenazer Wasti

October 22, 2026

AI can help employees complete tasks faster. But what happens to the rest of their jobs?

A company might introduce AI to draft reports, summarize information or handle routine requests. That can save time, but it does not automatically improve the broader workflow. Employees may still follow the same processes, wait for the same approvals and deal with the same bottlenecks.

This is where job redesign comes in. Instead of simply adding AI to existing roles, businesses need to rethink how work is divided, which decisions people make and what skills employees will need.

Research published by the CIPD in April 2026 highlights why this matters. Its survey of 1,342 HR professionals and business leaders found that only 46% of HR leaders were confident in their ability to lead job redesign, while 40% were confident in designing reskilling pathways.

The findings point to a gap between introducing AI and preparing people for the changes it brings.

For HR teams, the challenge is no longer just helping employees learn new tools. It is helping organizations rethink how work gets done.

What Is AI-Powered Job Redesign?

Job redesign means changing the tasks, responsibilities, skills, or decision-making authority associated with a role. AI can make this more important because it changes which parts of a job require human effort and which can be supported or handled by technology.

Consider a marketing team. AI might research a topic, summarize customer feedback, and prepare the first draft of a campaign. The marketing professional can then spend more time understanding the audience, developing the creative direction and deciding whether the content reflects the brand.

The role has not disappeared. The balance of work has changed.

This is the difference between task automation and job redesign. Automation focuses on making an individual task faster or easier. Job redesign looks at how that task fits into the wider role and what employees should do with the time, information and capabilities AI makes available.

If businesses do not rethink the wider process, they risk adding another tool without addressing the problems that slow their teams down.

What Does the Research Tell Us?

The CIPD’s 2026 research suggests that organizations may be more comfortable exploring AI than managing the workforce changes that follow.

The survey found that HR professionals were relatively confident in learning to use AI tools, with 67% saying they were confident in this area. Yet confidence fell when respondents considered the wider organizational changes AI requires. Only 40% of HR leaders were confident in designing reskilling pathways, and 46% were confident in leading job redesign.

These figures are based on respondents’ self-assessments, rather than direct tests of their abilities. Even so, they highlight an important challenge: knowing how to use AI is not the same as knowing how to reorganize work around it.

A separate Boston Consulting Group study published in September 2026 examined AI-leading companies through interviews with leaders across more than 50 businesses.

BCG identified five areas where these companies were changing how work gets done: the purpose of work, the tasks people perform, decision-making authority, team coordination, and how employees are rewarded.

The common thread was that these businesses were looking beyond individual AI tools. They were reconsidering responsibilities, workflows and the way employees contribute to business outcomes.

The study is qualitative, so its findings should not be treated as a universal formula for every organization. However, it offers useful examples of how businesses are putting job redesign into practice.

A Real-World Case Study: How Reserv Changed the Way Expertise Is Used

One of the more practical examples in BCG’s research comes from Reserv, an insurance technology company.

As described in the BCG study, Reserv developed a platform that allows employees to create AI agents for their work. These agents can capture parts of employees’ expertise and make that knowledge reusable across the business.

This approach addresses a familiar organizational problem: important knowledge often sits with a small number of experienced employees. When others need that knowledge, they may have to ask the same experts for help, wait for their availability, or learn through repeated trial and error.

Reserv’s approach helps turn some of that individual expertise into a resource that other employees and AI agents can use.

BCG reported two notable results from the company’s experience:

Learning and development programs: The time needed to build a new program fell from 18 months to four weeks.
Responses to regulatory changes: The time needed to respond fell from around three days to approximately one hour.

BCG also reported that junior employees could take on higher-stakes work earlier in their careers than they normally would.

These results illustrate how AI can change more than the speed of an individual task. When expertise is captured and made reusable, it can change how knowledge is shared, how work is allocated, and how quickly employees can take on new responsibilities.

There is an important distinction, though. The reported figures come from BCG’s qualitative study and company interviews. They are not a controlled experiment, and the article does not establish that the same results would occur in every organization.

Still, the example offers a useful lesson for HR leaders: job redesign can involve changing how expertise moves through a business, not just automating repetitive tasks.

For an HR team, the equivalent might be capturing approved policy guidance, standard operating procedures or experienced employees’ knowledge so that staff can find answers more easily. Employees would still need to review sensitive cases, interpret complex situations, and make decisions that require human judgment.

How AI Could Change Job Responsibilities

The Reserv example highlights several changes that HR teams may need to consider when redesigning roles.

1. Employees may spend less time on repetitive work

AI can assist with tasks such as summarizing documents, preparing first drafts, organizing information and answering routine questions.

The aim should not be to automate everything possible. HR should first establish whether a task is suitable for AI, whether the information is reliable and whether automating it will improve the overall process.

2. Reviewing AI output becomes part of the job

When AI prepares a report or recommends an action, someone still needs to check whether the result is accurate, appropriate and useful.

That means employees may spend less time producing a first draft and more time checking evidence, spotting errors and deciding what should happen next.

The level of review should depend on the risk. A routine internal summary does not require the same checks as a decision that could affect an employee’s pay, career or employment.

3. Skills requirements may change

Employees do not all need to become AI engineers. They do, however, need to understand when AI is useful, how to assess its output and when to rely on human expertise.

Depending on the role, this may involve stronger critical thinking, data literacy, communication, problem-solving and the ability to manage AI-assisted workflows.

4. Roles may become more connected

When AI reduces the time needed for individual tasks, teams may be able to handle work differently. Employees might take greater ownership of an outcome instead of passing each step to another department.

But this requires more than giving people access to AI. Managers need to clarify responsibilities, decision-making authority, and how employees will be held accountable.

How HR Teams Can Redesign Jobs for AI

HR does not need to redesign every role at once. A focused pilot can help an organization understand what works before making wider changes.

Start with tasks, not job titles

Choose a role or process where there is a clear problem, such as repetitive administration, lengthy handovers or delays in finding information.

Map the tasks involved, how much effort they require, where errors occur, and which steps depend on human judgment.

This gives HR a clearer starting point than asking whether an entire job can be automated.

Decide what AI should do

A practical starting framework is to divide tasks into three categories:

Automate: Repetitive, rules-based tasks with clear requirements and appropriate safeguards.
Augment: Tasks where AI can prepare drafts, summarize information or suggest options, while a person reviews the result.
Keep human-led: Tasks involving sensitive employee circumstances, complex interpersonal issues or high-impact decisions that require careful human judgment.

These categories are starting points, not fixed rules. The right approach depends on the task, the quality of the available information and the potential consequences of an error.

Connect AI to reliable information

AI tools are more useful when they can access the information employees actually need, such as approved policies, process documents and relevant business records.

Depending on the use case, this may involve a language model connected to an approved knowledge base through retrieval, an API, or an agent that can carry out defined steps.

Access controls matter. Employees should receive only the information they are authorized to see, and the system should not expose confidential employee data unnecessarily.

Redesign the workflow around people and AI

Once the suitable tasks are identified, map how the work should move from start to finish.

Decide which steps AI can handle, where employees must review the output, which situations require escalation, and who is accountable for the final result.

Involve the people doing the job. They often understand the practical problems, exceptions, and informal workarounds that are missing from processed documents.

Train employees and measure the results

Training should reflect the actual work employees will be doing, rather than focusing only on how to operate a new tool.

Start with a small pilot and measure outcomes such as time saved, error rates, response times, work quality and employee feedback. Also check whether the change has simply shifted work to another team or created additional review requirements.

If the pilot improves the process without creating unacceptable risks or workloads elsewhere, the organization can consider expanding it.

A Practical Example: Redesigning an HR Administrator’s Role

Imagine an HR team that spends a large part of its day answering routine questions about leave policies and preparing standard documents.

The organization introduces an AI assistant connected to its approved HR policies and document templates. Employees can use it to find relevant guidance, while the assistant can prepare first drafts of routine documents.

The HR administrator’s role changes in several ways. Instead of answering every standard question manually, the administrator reviews AI-generated responses, handles exceptions, checks documents that need additional care, and supports employees with more complex concerns.

Before the change, the team might measure how many requests it handles and how long each one takes. After introducing AI, it could also measure response times, answer accuracy, the number of cases requiring correction, employee satisfaction, and the time available for more complex HR work.

This is an illustrative scenario, not a reported company case. It shows how an organization could test job redesign using a specific workflow and measurable outcomes.

It also highlights why human oversight matters. AI should not independently resolve sensitive employee issues simply because it can produce a plausible answer.

The Opportunity Lens

AI adoption is often discussed in terms of tools, automation and productivity. But the bigger question is how businesses will organize work as AI capabilities develop.

The research from CIPD highlights a gap in confidence around job redesign and reskilling. BCG’s interviews offer examples of companies changing how they use expertise, distribute responsibilities and evaluate work.

Neither source suggests that every role should be redesigned in the same way. The right changes depend on the organization, its people, the work involved and the risks it needs to manage.

For HR teams, the opportunity is to help businesses make these decisions deliberately. Start with a real problem, understand the tasks involved, decide where AI adds value and give employees the skills and authority to work effectively in the new process.

The advantage will not come from introducing AI into every job. It will come from redesigning work so that technology and people can contribute where they are most useful.