Main content

HR, meet AI - everything changes...perhaps

Cath Everett Profile picture for user catheverett April 3, 2024
HR professionals are the guardians of both sensitive employee data and workplace policies and procedures. As a result, it is incumbent upon them to become a guiding light in terms of AI adoption not only in their own function, but across the wider business too, experts believe.

Human resources
(via Pixabay )

HR professionals are the guardians of both sensitive employee data and workplace policies and procedures. As a result, it is incumbent upon them to become a guiding light in terms of AI adoption not only in their own function but across the wider business too, experts believe.

A report by technology analyst firm Valoir entitled Is HR ready for AI? summarizes the situation nicely:

HR departments need to navigate the risks and opportunities of AI for their own HR processes and data as well as establish new policies and procedures for all employees’ use of AI and generative AI. This means having practical knowledge from a technology perspective. It also requires the ability to understand the human-related risks associated with AI, and to translate these into policies, practices and training that maximize AI’s benefit within and beyond the walls of HR, while minimizing its risks in a rapidly evolving technology landscape.

Fortunately, it would appear many HR professionals are trying to get a handle on generative AI at least. By mid-2023, more than three-quarters of those questioned had experimented with some form of the technology, the Valoir report indicated. 

More formal adoption rates are rather more mixed though, although they are rising. According to a Gartner survey conducted among HR leaders in January, 48% indicated they were in the process of evaluating such solutions.

A further 38% said they were piloting, planning for implementation, or had already deployed them. These figures have doubled since June last year due to both pressure from the top and interest from the bottom, sparked off by the buzz around ChatGPT. 

Eser Rizaoglu is Senior Analyst in the Gartner HR Technology Strategy and Management team. He points out that generative AI’s biggest use case today is in HR service delivery using employee-facing chatbots (43%).  Next comes HR operations (42%), where the technology is being deployed to generate policies and documents. Third is recruitment (41%), where is it being used to create job descriptions and analyse skills data.

AI adoption in the HR function

But, as Rizaoglu indicates, generative AI still tends to be deployed for individual, specific tasks rather than more broadly. So, for instance, while using it to write job descriptions may be becoming more common: 

It’s only a small sliver of the wider recruitment process. When data accuracy improves, hallucinations reduce, and it becomes clearer how to integrate large language models with the broader technology stack, it will then be used at more touch points and in more process steps.

Interestingly though, Rizaoglu also points out that HR functions are currently behind the curve compared to implementation levels in other departments. A survey of 822 functional leaders last November revealed that marketing tops the adoption list here. HR, on the other hand, was near the bottom due to professionals’ concerns over data accuracy and privacy. Once these key issues are sorted out though, he believes there will be a “greater proliferation of AI’s use within HR”.

Kate Obi, Chief People Officer at sector-focused Software-as-a-Service provider Advanced, explains the current hesitancy:

HR as a function tends to be slower at technology adoption than others due to its important role is helping to manage risk and ethics. It’s also one of the least data-driven functions, so there’s an element of needing to develop a better understanding of what to do with data.

But according to the Valoir report, AI offers a significant automation opportunity for HR:

Thirty-five percent of the HR employee’s workday is ripe for automation, meaning it is spent on tasks that can readily be automated in all or in part by AI (such as responding to emails or entering data). HR workers have already shown a high propensity for adoption of automation. The average HR employee already automated nearly 20% of their work tasks in the past two years.

The study pointed out that HR could most benefit from the technology in three key areas: recruitment, learning and development, and talent management. It also revealed that nearly 25% of organizations have already introduced generative AI for hiring purposes, making it the top area for adoption to date. A further 30% also plan to implement it in the next 24 months.

The importance of ethical AI

But on the downside, these three areas also pose the biggest potential risk of things going wrong. Concerns here include potential bias in AI-generated recommendations or using models unintentionally trained on data sets that reinforce bias.

As a result, setting up a cross-functional and diverse AI governance and ethics council is crucial, Obi says. Its role is to establish how AI should be used within the organization, to evaluate potential use cases, understand and navigate unintended consequences and work out how to achieve the best outcomes. As Rizaoglu points out:

HR leaders should play a big role here. They need to think about how AI ethical principles are managed in line with the risks. As the steward of employee experience, they also need to think about the implications both for the workforce and the HR function. 

But when Gartner asked HR leaders in January what their most common AI initiatives were, formulating ethical principles was only halfway down the list. Shorter term actions, such as training and up-skilling, developing internal usage policies and guidance, and devising an employee communications strategy, were given much higher prominence. But even work here left something to be desired, according to the Valoir report:

Even those that have adopted some AI have a long way to go in building and implementing policies and training that balance the necessary guardrails and appropriate flexibility. 

In fact, only 16% of employers had developed a policy for the use of generative AI, and even fewer for how to use it ethically. But this situation is already having consequences. 

A need for up-skilling

Uma Gunasilan is Professor of Management Development and Associate Dean of Research at the Hult International Business School. She refers to at least three cases she knows of where employees have started complaint and grievance procedures after their work, developed with the help of generative AI, was treated with scepticism – despite leaders having said in the past it was permissible to use such tools. As Gunasilan says:

These issues are already entering the HR realm. In most organizations, there’s no rule book, so it’s pushed to HR who bear the brunt of the situation. But many are saying they don’t know how to write policies for this or what they should be investigating in a grievance procedure or how to speak to managers about it. It’s a very real problem.

In terms of training policies, meanwhile, a mere 14% of companies had anything in place to help employees use the technology effectively, according to the Valoir study. Just eight percent had also introduced a development program for those whose roles could be replaced by it.

A key challenge at all levels, believes Gunasilan, is that HR professionals today simply do not have the skills themselves to deal with such issues. This is also borne out by Valoir’s report. It indicated that key barriers to AI adoption within the function include a lack of skills and expertise (26%), a fear of risk or compliance issues (23%) and a lack of resources or funding (22%). 

In skills terms, for example, it is imperative that HR professionals become more tech-savvy and learn how to craft, interrogate and analyse data more effectively, believes Obi. Critical thinking will also become more important over time as will mastering concepts, such as design thinking, and adopting more agile ways of working. She explains:

HR won’t have to be IT professionals but they will need to be commercially-minded people who understand governance, ethics, organizational dynamics and data so they can grasp how things flow system to system, or process to process. It’s about comprehending how things work to understand unintended consequences and make changes if necessary. So, the role becomes quite different.

How AI is changing HR

Rizaoglu agrees:

HR will have to focus on the underlying practices that enable technology, which means shifting from a view that tech is the solution rather than an enabler. The focus is on data literacy and dexterity. In more mature organizations, we’re also seeing HR functions shift from being more process-led to being more experience-led. This involves thinking more about the employee experience and treating people more like customers. So, ‘what are the journey maps, what experience do we want them to feel, and what systems can help us achieve that?’ rather than ‘this is the process and the experience you end up having as a result’.

While such a shift in focus is not necessarily due to AI alone, “it may help achieve it”, Rizaoglu adds. The upshot is that HR headcount is unlikely to fall much. Instead, AI will likely be used to automate transactional activities, leading to the creation of “augmented roles”. He explains:

One of the most frequently cited use cases is that of HR chatbots. But even though a large proportion of employee services will be managed this way, headcount will reduce by less than five percent. This shift to more of an experience and product-led approach will mean that HR has to build up the necessary skills though or go to other functions, such as IT, customer service and marketing, to teach them

Although this approach is uncommon today, Rizaoglu has already seen some companies reorganize the HR function to mirror the employee journey. So, rather than placing HR business partners in different business functions, professionals focus on different stages of the employee lifecycle, ranging from attraction and selection to end of life. He expects a “reasonable number” of companies to have gone down this route by 2030. 

But such a scenario also means that HR will become much more “infused” within the business, believes Gunasilan:

We’ll still have HR departments, but HR professionals will work very closely to understand certain teams or liaise more closely with managers. This will help in data-driven decision-making, enhancing wellbeing, strategic workforce planning, mapping competencies and responding to changing market demands. HR functions are going to be highly strategic. They’ll be a guiding beacon and lighthouse to all the other teams who don’t know how to do this, which means it’s vital they get on top of all of this themselves.

My take

The message here is clear. It is the role of HR professionals to light the way in terms of ethics, risk management and up-skilling the wider organization. But they really need to up-skill themselves and get a good handle on how to deal with both data and AI if they are to do so effectively.




A grey colored placeholder image