As AI agents move beyond experimentation, organizations face a more complex question than where to deploy the technology: How should work itself change?
A new report from The Conference Board, produced in collaboration with SkillsRight, examines the strategic choices leaders—particularly CHROs—need to consider as AI changes how work is organized, including how work is divided between people and AI, what skills redesigned roles require, and how workforce and business value are measured.
The research draws on interviews, focus groups and hands-on research with senior HR, talent and AI leaders.
The report comes as AI investment continues to accelerate. According to The Conference Board C-Suite Outlook 2026, 43.6% of executives identified AI and technology as an investment priority, more than any other investment choice.
But the report makes clear that capturing value requires more than buying tools or launching pilots. Organizations also need to determine the business outcomes they want to achieve, decide how work should be divided between people and AI, establish governance, measure results and determine how any gains will be used.
A Five-Stage Framework for Redesigning Work Around AI
The report outlines five stages:
1. Start with the desired business outcome.
Organizations should first identify the outcome they want to improve, then determine where AI agents can add value. The report cautions against deploying agents simply because the technology is available.
2. Redesign and allocate work before deploying agents.
Leaders should break workflows down task by task to determine what AI should handle, what people should handle and where they should work together.
The report recommends considering factors including quality, judgment, accountability, the consequences of errors, employee impacts and customer value. People should remain accountable for results, even when agents perform work autonomously.
Organizations should also deliberately redesign the human role, including the skills, training, job level and potentially compensation required for the work that remains.
3. Build the infrastructure, governance and budget agents require.
Deploying AI agents requires more than the technology itself. Leaders also need to account for data, permissions, controls, monitoring, maintenance and human supervision.
The report notes that the full cost of operating an agent can extend well beyond vendor or model fees. In one illustrative model, direct AI-model use represented just 9.4% of an agent’s recurring monthly cost.
4. Measure what people and agents produce together.
Organizations should measure whether redesigned work improves quality, speed, cost or other business outcomes, rather than simply whether an agent completed a task.
The report also cautions against automatically counting time freed by AI as a financial gain. Organizations should first demonstrate how that time was used and what measurable result it produced.
It recommends distinguishing among three types of gains:
- Cash savings: Spending that actually disappears, such as overtime, contractor costs or other expenses.
- Avoided future costs: Planned spending or hiring that is no longer required.
- Higher-value work: Employee time redirected to other work that produces a measurable result.
5. Decide where the gains go.
Verified gains can be used to reduce costs, help organizations handle more work without equivalent hiring, support reinvestment or deliver benefits to employees and customers.
The report also recommends clearly communicating how those gains will affect jobs, staffing and employees.
Protecting Early-Career Talent Pipelines
The report also highlights an important workforce consideration: when AI takes over routine work, organizations risk removing the tasks through which employees have traditionally built judgment and experience.
To address that risk, leaders should deliberately replace those learning opportunities through approaches such as mentoring, rotations, supervised practice and progressively more complex work.
HR and IT Must Work Together
The report also underscores the need for cross-functional ownership. While IT may own the technology, HR should help shape roles, skills, staffing, training and change management wherever AI changes how work gets done.
Business leaders, technology teams, finance and HR each have a role to play in ensuring that AI deployment is connected to measurable outcomes and that the workforce evolves alongside the technology.
As organizations determine how AI agents will change work, the report emphasizes the importance of defining desired business outcomes, determining how work should be divided between people and AI, and ensuring roles, skills, training and change management evolve along with the technology.
For SkillsRight, the research connects directly to a central question in skills-first workforce transformation: how CHROs and other leaders use skills to redesign work as AI capabilities evolve. The report shows where skills practices and AI deployment converge at the task level, where leaders identify the skills required and consider how work might be allocated across AI, people working with AI, and people alone.
Read the full report, A Framework for Agentic AI and Work Redesign, available for public download on The Conference Board website. https://lnkd.in/eCBm7nbw