A Review of the First Half of 2026 in HR Consulting

At the halfway point

Looking back from the midpoint of 2026, one word captures the first half of the year in HR consulting: transformation. Economic uncertainty made companies cautious about consulting budgets overall, but demand related to AI grew markedly. The nature of that demand changed as well. A few years ago, clients asked what HR practices they should adopt or how to build a system. This year, the question at the consulting table has been: “How can AI actually change our HR function and the way we work, and where should we begin?”

A change in the question changes the work of consulting. Three developments in the first half show how the market is evolving and what companies should focus on next.

First, the question changed from DX to AX

For more than a decade, digital transformation in HR has centered on implementing systems and organizing data. But this approach had a structural weakness: it relied on employees and managers to enter data manually. When they treated data entry as a formality, data quality suffered; poor data then eroded trust in the system. Many companies experienced this cycle.

AI transformation, or AX, starts where DX reached its limits. It goes beyond automation or the adoption of a new technology to address gaps that earlier efforts could not, raising the value of HR work itself. AX depends on context that is less structured, more conceptual, and often more sensitive than policy documents or standardized data. What makes a goal a good one? What does performance mean in this organization? How are jobs and skills defined? AI must be able to work with this context to be useful in HR. The first half reinforced a central lesson: domain expertise is essential to successful AX.

This challenge is not new. In the DX era, some large Korean companies adopted global software packages only to reverse course years later because they could not bridge the gap between standardized systems and the realities of Korean HR operations, group-company policies, and organizational culture. If differences in policies and culture were the customization hurdle in the first wave of DX, domain expertise grounded in each company’s HR philosophy and operating context is the key to customization in an AI-enabled second wave. Standard technology can be purchased; AI that understands an organization’s context must be built.

The way demand for AX consulting emerged was equally revealing. Executives in many companies were drawn to the prospect of running leaner workforces with AI, while practitioners recognized that AI could not simply replace entire jobs. That gap in expectations generated consulting work: AI pilots and validation, changes to roles and responsibilities as structures and jobs evolve, and assessments of workforce impact. At this transitional stage, companies are still defining what to apply, where, how, and to what end. Helping them define that direction has become a new role for consulting.

Second, disappointment with AI features drove interest in agents

The second development was a frank reassessment of first-generation AI features. Over the past year or two, HR products have added buttons that suggest draft goals, summarize review comments, or answer simple chatbot questions. Users have been underwhelmed. Companies are no longer impressed by an AI tool that lies dormant most of the year, proposes a few goals in January, and summarizes one review in December.

The underlying problem is a design assumption, not simply the technology. Those features were attached to separate stages of a process still run entirely by people. They work in isolation, lack the wider context, and often appear only during a particular season. The market is beginning to ask for an agentic approach. The full performance management cycle, from setting goals and monitoring progress to feedback, one-on-ones, assessments, and reviews, must become context an agent can draw on. Only then can it, for example, flag a mismatch between a rating and the record of work during a calibration session, or detect signals of recency bias or overly generous ratings.

This connects to the longstanding goal of continuous performance management. A company closes its books at year-end, but accounting happens throughout the year. Likewise, performance management needs both periodic reviews and an ongoing flow of progress checks. Sustaining that flow through human effort alone has been impractical, so performance management has often become a largely formal exercise centered on the review season. Agents may make continuous management practical for the first time. In a fast-changing environment, a goal set in January may not remain relevant in December. An agent can monitor data throughout the year and proactively suggest a goal adjustment or a conversation.

Agents are changing the design of the user experience, too. Much of the AI in global HR products still appears as an assistant or copilot attached to a particular screen, but possibilities are expanding beyond the chat window. An agent might brief a manager on team members needing attention through a card on a morning dashboard, suggest language inline while feedback is being written, or remain invisible until a timely notification is needed, an approach sometimes called “zero UI.” How and when AI appears should depend on a person’s role and situation. A tool that merely looks like another chatbot will struggle to earn employees’ trust or executives’ investment.

The first half also exposed two needs for balance. One is to avoid excessive personalization. Goals and assessments are inherently organizational: they need alignment with higher-level objectives and fairness among peers. An AI confined to a personal assistant’s perspective will have only part of the picture. The other is data. Much of the evidence of work lives outside HR systems, in collaboration tools, email, and project management platforms. AI built only on HR system data may give thinner answers than a general-purpose model. Connecting data across systems and governing its use must therefore become a core part of HR’s AI agenda.

Third, consulting itself became a target of transformation

AI is changing the consulting industry as well as its subject matter. In global markets, the shift is becoming structural. AI is taking on substantial portions of the research and presentation work once assigned to junior consultants, challenging the traditional pyramid staffing model. Some major firms have formally moved toward fees linked to outcomes rather than hours worked; one of the world’s largest strategy firms has publicly said that roughly a quarter of its fees are performance based. Alliances with AI vendors are also supporting a delivery model in which engineers work on site with clients and take responsibility for implementation. While the traditional market for HR policy advice has been relatively stagnant, AI and AX consulting, though still smaller in absolute terms, has become one of the fastest-growing areas. Similar signals, including reduced demand for research assistants, can be seen in Korea.

What interests me most is the change in deliverables. The era in which consulting ends with a report is fading. Consulting knowledge accumulated over many years, such as the characteristics of a good goal, standards for job-specific capabilities, and criteria for evaluation and feedback, can now be embedded in systems as knowledge layers and instructions that AI consults continuously. A common foundation of domain knowledge can be combined with each client’s context. In effect, domain expertise is becoming a product. An end-to-end model that connects diagnosis, design, implementation, and adoption is emerging as a new standard. That creates an opening for specialized firms with both deep domain knowledge and the means to execute, even when they are smaller than the major consultancies.

Three developments to watch in the second half

I expect three developments to shape the HR consulting market in the second half of 2026.

First, organizations will move from proofs of concept to deployment. If the first half was about testing possibilities, the second half will bring more agents into defined HR domains such as performance management, payroll, and talent search. Vendor-neutral agent architecture and the orchestration of specialist agents will become new advisory needs. Agents embedded in systems and working across their data will coexist with external agents that answer natural-language questions. General-purpose AI may even connect to an organization’s HR systems to ask whether a proposed goal aligns with its direction. HR technology architecture will need to become more agent ready.

Second, context governance will move to the center of the agenda. Differences in AI performance increasingly depend on the quality of an organization’s contextual data, not just on the model. Turning tacit organizational knowledge, including job and skill frameworks, goals and assessment criteria, and operating principles, into forms AI can read will become a recurring project. This is, in effect, the work of building an HR ontology.

Third, consulting will become more accessible as a service. A new layer of AI-enabled “micro-consulting” is emerging between full-scale projects and self-service tools. A midsize or smaller company with limited HR infrastructure can use AI guidance to establish the foundations of its policies, bringing in an expert where judgment is needed. This could materially broaden access to consulting.

Three priorities for companies

Against this backdrop, I would suggest three priorities for HR leaders.

First, make context an organizational asset. Successful AI adoption depends less on which tool is selected than on how clearly a company has defined work and performance. Organizing its own job, skill, goal, and evaluation frameworks as usable data is the starting point for AX.

Second, redesign processes instead of attaching features. Adding AI to existing steps has shown its limits. Processes should be designed with the assumption that agents are available throughout the organization. At the same time, the redesign must respect the real calendar of HR decisions, including reviews, promotions, and assignments. A phased approach should begin where business impact is clear.

Third, design how people and AI work together. AX should relieve people of complex, unproductive tasks and data entry that degrades data quality, freeing them to focus on judgment and conversation. Organizations need to define how human employees and digital workers collaborate, then invest in change management and AI literacy. That is how technology becomes performance.

Looking ahead

In the first half, the market began to see how AI could change HR. The second half must demonstrate that direction through implementation. The technology has arrived quickly. The remaining work is for each organization to organize its context, redesign its processes, and establish ways for people and AI to work together. HR and HR consulting will be at the forefront of that work in the second half of 2026.

HCG's AX Consulting is built to carry this work through: a five-stage approach, Strategy, Design, Build, Run, and Enablement, connects each diagnosis directly to execution by linking the resulting design to HCG's own AI HR solutions, including elizax, hunel, JaDE, and talenx.