Not anti-AI. Pro-reality.

Let’s get that straight right at the start: I am not anti-AI, I am pro-reality. AI will absolutely have major impact on HR Tech — but on its path, I see a major risk that many AI pilots will not provide the expected AI-based transformation of HR operating models. There is a chance, that powerful AI for HR initiatives will follow Gartner’s Hype Cycle and vanish into the trough of disillusionment for years before unleashing its true potential. Hundreds of proof-of-concepts will outline the potential of AI in HR but not transform the HR operating model fundamentally and therefore become some punctual productivity boosters at maximum.

In March 2026 Lakhani, Spataro, Stave published research insights explaining why “the last mile problem slowing AI transformation”. They provide a pattern which applies also in general in HR Tech and I do see further particular reasons why AI will take longer to fundamentally change HR Tech than the hype suggests. Some of them are not problems, they are good reasons why HR should set its own pace for an AI transformation. My intention here is not to slow down AI innovation in HR. I want to explain why the pace of AI adoption in HR Tech is not the one we observe in other business functions. And not all of the reasons I will make the case for apply equally to every HR Tech solution, every HR organization, or every vendor — but they represent the typical situations and realities I observe today.

So here are five frictions I observe that prevent AI to change HR Tech fundamentally in the short term.

1. The HR organization and infrastructure isn’t ready yet

Since I am active in the area of HR Tech the implementation of technological innovations faces similar challenges: Unclear expected outcomes (“what to achieve with the new technology”) combined with unfinished conceptual, technical and organizational homework to set the base for leverage the innovation. Data cleansing, process redesign, clarification of governance, and proper change management – to name some main topics. Gaps in fulfilling these critical prerequisites led to small pilots with limited features — buying time to prepare the organization for full adoption. And still, many companies are stuck in such a journey. Today I see organizations with severe data quality issues, ongoing disputes around process standardization, and disruptions caused by unclear ownership and governance. Implementing AI on top of these unresolved foundations will not fix them — it will amplify them. Poor data produces wrong results, and wrong AI results damage trust. Unclear governance leads organizations to limit AI just to retain control, and in doing so, they lose the very power AI could provide.

The AI vendor landscape in HR Tech adds another layer of complexity. It is fragmented, and in some cases, AI is little more than a marketing label attached to features that existed long before. All major HR Tech vendors and many start-ups make bold promises and present compelling cases for their own AI technologies. But most of these solutions have not been battle-tested at scale. We do not yet know whether they deliver the expected value or justify the ROI expectations — we are still in an early phase of this market. This could provide reasons why HR departments might be reluctant to start a powerful and comprehensive AI transformation and rather sit and wait, especially in the current cost-efficiency driven economic situation where budgets for innovations are tight.

A third infrastructural blocker is one that has slowed HR Tech innovation before: the unresolved challenge of comprehensive cross-application integration. A typical HRIT landscape consists of a mixture of 10 to 30 global solutions — and sometimes hundreds of local ones — all with very different origins. From cutting-edge platforms to legacy systems still running on outdated architecture. From large vendor suites to small custom-built tools. From globally standardized, internationally capable systems to highly localized, country-specific applications. Creating consistent workflows or meaningful analytics across such a fragmented landscape is a major challenge for every HRIT organization. There is no reason to believe AI will escape this same roadblock. A powerful HR AI would need access to most of these systems to connect the dots and materialize the potential that has so far been out of reach. Leaving aside whether organizations are willing to grant AI that level of access — most will not even be technically capable of providing the APIs and integration capabilities required to make it work.

2. The budget is stuck elsewhere

SAP announced the end of mainstream maintenance for SAP HCM on-premise years ago and has since offered multiple transition paths to support the move of payroll to the cloud. Yet many companies have not started that journey — or are only at the very beginning. With the extensive customization built into local payroll solutions over the past decades, this transition is anything but a simple lift-and-shift. In many organizations in Europe, this is priority number one for the foreseeable future, and it consumes a significant share of HR Tech budgets. Because it involves payroll and because the running solution has a defined end date, the topic is both important and urgent. Innovative HR Tech leaders feel this acutely: the transition will come with limited room for innovation and will continue to dictate HRIT roadmaps and budget plans for the next one to three years.

Over the past 10 to 15 years, larger companies around the world made substantial investments in SaaS platforms for HR — implementing global solutions like Oracle HCM, SAP SuccessFactors and Workday, but also buying from regional and local vendors. These platforms condense established HR data models and process designs into one integrated solution, with the goal of supporting the entire employee lifecycle in a user-centric environment. The organizational change this required was enormous. Beyond the cost of technical implementation, organizations invested heavily in adapting their HR service delivery to the new technical reality. It will be a hard case for AI vendors to make when their proposition targets functions, features and processes already covered by these platforms. Those platforms are live, and companies spent significant money to make them work. It is difficult to imagine a board — and especially a CFO — accepting the sunk costs involved if the business case suggested replacing a running SaaS platform with AI technology. Despite the complaints HR Tech leaders regularly hear about their HR platforms, these solutions generally run in a stable and scalable way. Why would any organization give that up?

How hard this case is to make becomes clear when you look back at the time when vendors offered new technologies to move HR services to the cloud. In those days, innovative HR Tech leaders struggled to build financial arguments for that step. Although the new technology brought genuine innovations — particularly in the areas of Employee Self-Service and Manager Self-Service — the legacy system was live, running and fully amortized. A direct and immediate ROI for the new solution was difficult to prove, and as a result, the decision process for moving HR admin into a SaaS solution took years in many larger companies. AI technologies might fall into the same trap when they are positioned as a replacement for existing and functioning HR SaaS solutions.

But what about the argument that AI could drive efficiency in areas the SaaS platforms do not yet cover? In some areas of HR — Talent Management or Succession Planning, for example — efficiency is not the primary value driver. But in HR Services it is, and that deserves a closer look. HR Services operate in a strictly rule-based environment, governed by labor law, tax regulations, and company agreements with works councils and unions. In this context, automation is already available, established and good enough to close the most significant efficiency gaps. Organizations with mature HR Services have invested substantially in automation over the years. The effort required to close any remaining gaps with AI technology would rarely justify the business case — both because of the investment required to implement and adapt the new technology relative to the marginal efficiency gains still available, and because of the rebound effect: the unforeseen behavioral and systemic responses that follow realized efficiency gains. Beyond that, stable and scalable automation is simply the better fit for most HR Services processes. What is needed here is not a technology that analyses patterns and makes interpretative decisions — it is reliable, repeatable and rule-based handling of well-defined processes.

3. The human factor is underestimated

People working in HR departments typically make a conscious decision to spend their professional lives in an area that is human-centric and employee-focused. In the past, this mindset sometimes made it difficult to bring everyone on board for HR Tech programs and projects. “I want to work with people; I did not join HR to become an IT expert ” is a statement I heard more than once in my career. Part of my job as an HR Tech advisor has also been to convince resistant HR professionals that getting closer to HR Tech is not a betrayal of their values — it is a prerequisite for keeping HR service quality high and meeting the expectations of employees who are used to digital services in every other area of their lives. And yet, those HR experts had a point. There are moments where the human touch of a manager or an HR partner matters more to an employee than any digital solution could. The ambition has always been to balance the advantages of new technology with the human need to interact with other people — especially in significant moments, whether positive or difficult. When it comes to integrating AI into HR, vendors need to address this tension directly: how will AI support HR experts in those high-touch moments, rather than turning the organization into a technocratic machine?

What I currently observe in the market is an emerging focus on how AI can take over specific HR tasks and help reduce HR headcount. Benchmarks are available showing the FTE impact of AI adoption, and vendors run webinars helping HR leaders calculate personnel cost reductions from implementing their solutions. This kind of narrative was also present in the early days of SaaS in HR. Investment cases were built on significant cost reduction through freeing up HR FTEs, and that message spread quickly through HR organizations. It made HR Tech implementation programs extremely difficult to manage — especially when those programs depended on the functional expertise of the very people who feared being cut. Some responded by surfacing endless new complexities and reasons why the new solution could not cover all functional requirements. Others convinced their local business managers that there were good reasons to keep their branch out of scope. The positive momentum for investing in new technology was gradually consumed by the fear of a cost-cutting program operating behind the scenes.

There is a further challenge familiar from technology implementations in the past: the blame-the-system fallacy. A lack of clear governance and process definitions — combined with an unwillingness to adapt to the new solution — led HR teams to blame the technology rather than the underlying organizational decisions. The new platform provided transparency on adherence to HR standards across the organization and helped HR move away from outdated practices. But communicating that message fell to local HR contact persons, who were often expected to explain a new service delivery reality to their internal clients. For some, that was a difficult message to deliver. Blaming the new system was simply easier than defending a management decision to standardize and streamline HR. Over the past ten years, I have encountered several organizations where the reputation of a new HR SaaS solution had suffered so badly that management decided to invest in a migration to a different platform. Interestingly, in most of these cases there was no evidence of poor vendor performance – it was the result of the blame-the-system fallacy.

If the critical ambassadors for AI technology in HR — the local HR experts — believe that AI will magically resolve issues rooted in unclear governance, poor data quality or insufficient process definitions, they will be disappointed. And if they fear that AI is coming for their role, they will start blaming it for the wrong reasons. Given that these individuals handle sensitive data and given that people are cautious about sharing personal information with AI systems, local HR ambassadors can have a decisive influence on whether local management and employees ever develop trust in the new technology.

4. AI in HR is real — but the narrative is running ahead of the reality

Reading articles, blogs, LinkedIn posts and conference agendas, one could easily get the impression that AI has already replaced most HR Tech solutions — or at least catapulted them to a fundamentally different level. Companies currently in the middle of SaaS transformations may feel that their existing investments are sinking into obsolescence. But there is significant potential for misreading what is actually happening in the HR Tech market.

First, the investment side: yes, substantial capital — from vendors and venture capital alike — is flowing into AI for HR, and meaningful innovations can be expected throughout 2026. But when it comes to HR Tech in European companies specifically, many organizations are still working their way towards the plateau of productivity with SaaS. Many transformation programs remain focused on moving the employee lifecycle to the cloud and establishing a future-proof tech stack for payroll. AI may well play a significant role in this in the future — but right now, the foundational homework on data quality and technical architecture is far from complete in many organizations. And AI does not produce reliable results on poor data. A substantial part of the market still needs time to stabilize and run HR Services robustly on the transformed landscape — and will therefore have neither the resources, the budget nor the bandwidth to pursue major AI programs in the near term.

From a top management perspective, there are also rational reasons not to position HR as an early adopter of AI. Yes, AI in HR can improve efficiency — but is that a C-level priority when HR typically represents only one to three percent of total operating costs in large companies? Compare the AI automation potential in HR with that of Manufacturing and Supply Chain, where processes are highly repetitive, rule-based and data-rich. Or Finance and Accounting, with its high volume of transactions, reconciliation and reporting. Or Sales and Marketing, where automation of personalization and pricing optimization is a clear competitive lever. Plotted on a simple matrix with AI automation potential on one axis and total cost impact on the other, HR, talent and organization would probably sit firmly in the lower-left quadrant — and therefore well outside the top management spotlight.

Research from Fosway Group confirms these patterns in the Learning Tech market: in 2025, only 10% of AI features tracked by Fosway were live with customers. By early 2026 that had risen to 24% — meaningful progress, but still far from the picture painted by vendor marketing. Critically, Fosway notes that a feature reported as live does not necessarily mean it is fully rolled out across the organization — only that it has been technically activated in the product.

So where does the urgency and hype actually come from? Part of the answer lies in the dynamics of a young and fast-moving market. Many organizations — from innovative start-ups to established consultancies — have built compelling narratives around the transformative potential of AI in HR, and have made significant commitments to bringing those visions to life. This creates a natural momentum: the conversation moves fast, use cases multiply, and the pressure to act feels real. It is worth keeping in mind, however, that market narratives in early-stage technology cycles tend to run ahead of enterprise adoption realities. Some of the use cases currently in the spotlight address challenges that existing solutions already handle well — which means the urgency to act may be lower than the noise suggests. As with any emerging technology, separating genuine innovation from market momentum requires a calm and independent read of the landscape.

A similar dynamic is visible on the vendor side. The major SaaS players are actively integrating AI capabilities into their platforms — and this is genuinely significant. Much of this work is currently focused on improving the technology stack: optimizing platform performance, refining configuration standards and incorporating leading practices at scale. These are meaningful developments that will over time translate into better products for customers. The pace and focus of this integration will vary across vendors, and it is still early days in terms of what customers will directly experience. Watching how this evolves will be one of the more interesting stories in HR Tech over the coming years.

5. HR is solving the wrong problem

Looking at today’s typical AI in HR pilots, three focus areas dominate: product and policy navigation — for example via chatbots supporting employees as tier-0 in HR Services; pattern recognition and data consolidation in complex environments such as skill management; and process automation for repetitive tasks, an area still in its early stages. Some voices in the market are already imagining a “dark HR factory” — a fully automated HR department requiring no human HR resources at all.

What these narratives share is a common driver: increasing HR efficiency. Even the skill management use cases, on closer inspection, follow this logic. In most cases, existing HR technology — without AI — is already capable of fulfilling the proposed use cases. It would simply require more resources and time to analyze patterns and extract deeper insights from talent data. Genuine added-value use cases for AI in HR are difficult to identify, and the majority of what is being discussed today would support HR Services — not HR Business Partners.

Latest research published in the Harvard Business Review adds an important dimension to this picture. Studies show that generative AI helps non-experts reach a solid level of competence in a given area more quickly — accelerating the upskilling curve for those earlier in their development. However, the effect on genuine experts is significantly smaller. This phenomenon is sometimes referred to as the AI wall: generative AI can bring non-experts up to a certain level, but navigating the challenges beyond that threshold still requires real expertise, contextual judgement and lived experience.

A further HBR study is worth considering here. Studies examining how large language models respond to strategic business challenges found that LLMs consistently gravitate towards recommendations aligned with current management trends and business buzzwords. Even when detailed and varying context was provided, the models returned similar advice — advice that did not resolve the specific business problem at hand, but rather reflected the dominant thinking of modern business culture. As one researcher put it, LLMs are not analyzing your specific business — they are offering a polished version of popular answers that sound good. Even more concerning, when faced with genuine binary business dilemmas, some LLMs showed a tendency to propose hybrid middle-ground solutions — responses that would leave organizations precisely where strategic clarity is most needed: stuck in the middle.

Taken together, this means that current AI technology has limited potential when it comes to supporting HR in becoming a genuine strategic business partner. The risk is that HR responds to this reality by doubling down on what AI does best today — driving efficiency — and building its case around the narrative of doing more with less. That path leads not to a stronger HR, but to a less visible one. Organizations that are seriously embracing AI across their operations do not need less HR — they need more of it. More HR expertise to navigate the boundaries between human work and AI agents. More capability to ensure that the augmentation between people and technology succeeds in a way that works for shareholders, leaders, employees and all other stakeholders. That is the problem HR should be focused on solving.

Key take-aways for HR

Let me be clear once more: I am not anti-AI, and I am not arguing that AI will have no impact on HR Tech. It will. My intention is not to provide arguments against starting, but arguments for starting differently. Here are five recommendations for navigating AI in HR initiatives with more clarity and confidence.

• Do your architectural and functional homework. New AI technologies carry enormous potential — but that potential depends entirely on the foundations beneath them. AI requires a stable technological basis, meaning solutions that are genuinely AI-ready, and clear functional goals defining what you actually want to achieve. AI will not create that foundation for you, but it can help you to orchestrate a complex landscape with that strong foundation. Organizations that skip this step will find that AI amplifies their existing problems rather than solving them.
• Start small and take an evolutionary approach. Major budgets are already committed to payroll transitions and SaaS programs — that is just the reality for most organizations right now. But this is not a reason to stand still. Small, well-chosen investments can prepare your organization and your HR users for an AI-enabled HR function. The argument that payroll and SaaS must come first should never become an argument for not starting at all.
• Focus on local and tangible use cases first. Too often, corporate AI initiatives underestimate the importance of local HR realities, and local complexities have a habit of limiting global solution approaches significantly. AI deserves a realistic chance to prove its value — and that means putting local HR at the center of your early initiatives, not the periphery. Start where the impact is concrete and the feedback is immediate.
• Set your own pace — but make sure you are moving. The urgency to implement AI in HR should come from your organization’s specific situation and needs, not from market trends. The risk of following hype blindly is real — but so is the risk of standing still while other organizations around you are learning. Both extremes lead to the same outcome: you arrive late, unprepared, and without the experience that only comes from doing. The goal is not to rush. It is not to wait. It is to move deliberately — small, well-governed steps that build organizational readiness, test real use cases, and keep your options open as the market matures. Organizations that start now with a focused and realistic scope will be significantly better positioned in two to three years than those that waited for the perfect moment. 

But most important in this early market phase as of early 2026: Understand how AI can support HR’s strategic role — not just its operational efficiency. The most important opportunity AI offers HR is not to do more with less. It is to help HR become a genuine leader in the intersection of human work and AI-enabled organizations. Labor market data in Germany show that the share of jobs requiring complex activities has grown significantly in the recent years – and AI will accelerate this growth. Position HR as the function that helps leadership teams understand how to leverage AI in a way that benefits all relevant stakeholders — and that includes employees and means top talents first and foremost.

There is also a broader opportunity here that HR has not yet fully claimed: HR has developed a complete portfolio to manage the entire life-cycle of the human workforce in the organization. At a specific maturity level, the organization will also require active management of the digital workforce:  onboarding the AI agents with company specific information, providing learning activities to develop them further or providing scalable assessment tools to help the leadership team rating their quality, potential and performance. HR developed all these concepts for the human workforce and there might be the opportunity for HR to translate those concepts for the digital workforce, too.

The organizations that get this right will not be the ones where AI experts tell HR what to do. They will be the ones where HR experts have a seat at the AI table.

Want to overcome the frictions slowing your AI-based HR transformation?

Contact HR Tech Navigator for:

✓ Knowledge infusion on AI-technology in HR 

✓ AI-Readiness assessment of the HR Department 

✓ AI-based HR Transformation Roadmap 

martin.wild@hrtechnavigator.com

hrtechnavigator.com

Sources:

Chui, Manyika, Miremadi et al for McKinsey & Company, April 2018: “Notes from the AI Frontier; Insights from hundreds of use cases”

Chui, Hazan, Roberts et al for McKinsey & Company, June 2023: “The economic potential of generative AI”

Fosway Group, March 2026, “Fosway AI Insights 2026: AI Market Assessment for Learning Systems”

Lakhani, Spataro, Stave in HBR.org, March 9, 2026: “The “Last Mile” Problem Slowing AI Transformation”

Romasanta, D.W. Thomas, Levina in HBR.org, March 16, 2026: “Researchers Asked LLMs for Strategic Advice. They Got “Trendslop” in Return.” 

Otter on Substack, February 18, 2026: “SaaS and the impending doom?”

Audrey Quick on LinkedIn, 02.09.2025: “Stop Blaming the System: The messy truth about bad workflows”

Vendraminelli et al. in Harvard Business Review March-April 2026: “Gen AI won´t make your employees experts”

Weber, Die ZEIT 11/2026: “Die KI ist kein Jobkiller!”