This article by Edward Houghton, Senior Research Advisor for Human Capital and Governance at the CIPD, takes a look at how automation is likely to impact HR and the role it will need to play in bringing ethics into a future filled with AI, data and algorithms.

In a future of AI, data and algorithms, HR’s ethical centre will be key. With much of the hype around technology in the workplace ramping up it can sometimes feel daunting to think about how AI or robotic process automation (RPA) is likely to change HR practice. The function has barely got used to using people data, let alone applying robotics and automation to its practices. Other professions outside of HR are also seeing their tasks be automated: IT in particular is one such function where automatable processes are now being handed over to robots.

It is useful to understand what kind of automation is happening to help us explore the kind of impact that might be expected in the future. At its most basic, RPA is based on repetition, following simple structured rules. Basic RPA works well on large volumes of well-structured data which indicates a clear beginning and end. No ambiguity. For example, an administrative task of copying user details and pasting them into online database forms can be automated - an event set to begin on the receipt of an email, or at a time when a database is updated, say first thing on a Monday morning. These tasks, whilst repetitive, are still prone to human error but with RPA, much of this error can be reduced.

More advanced than basic RPA are technologies which make use of natural language processing and machine learning. These systems are often designed to bring together both structured and unstructured data, often in large volumes, and can rapidly provide analysis of these data sets and apply them to practice. IBM Automation with Watson is one of the best known examples of this type of RPA, for example the use of chat-bot technology to deal with employee password resets, in which a password reset is actioned and delivered by an end-to-end automated process. This kind of cognitive computing is becoming more prevalent, and in the above example, often actioned by a human in an IT support centre, can now be fully automated. The potential is that this frees up the IT technician to tackle more cognitively challenging and complex tasks which at present cannot be automated. More complex and sophisticated RPA (sometimes termed intelligent RPA) are currently in development but as of yet have not been rolled out beyond the laboratory.

Organisations who are currently implementing this technology have started to from a position of efficiency: processes which currently take up large amounts of resource but which deliver incrementally small returns are being automated so that experts can be freed up to work on more important, and often strategic tasks. This poses some interesting dilemmas for HR in these organisation as to when and how automation of tasks is most appropriate: is it when repetitive tasks are overwhelming and no longer cost-effective to resource? Is it when functions are required to become more strategic in delivery but are burdened by operational processing? Or is it to de-risk processes which are prone to human error that, over time, can cause significant risk? These business or performance oriented outcomes are what dominates the decision making of most functions on this topic.

There is however an ethical view on automation which poses some interesting dilemmas for HR: namely if automation brings greater well-being and security to individuals, is this reason enough to automate even at high-cost or minimal productivity gain? In these scenarios, such as the automation of high-risk roles, certain tasks which pose a risk could be automated to improve health and safety outcomes or even broader well-being. High-risk high value tasks in oil and gas industries are a good example of where it is both cost effective and positive for health and safety and well-being for tasks (e.g. remote working) to be automated, therefore improving outcomes for employees.

A key issue however arises when we consider the unintended consequences of the large-scale automation of tasks: the issues arising from ethics. At present it appears we aren’t doing enough to explore some of the ethical tensions that arise when we automate tasks or jobs. It’s sometimes useful to consider a workplace whereby the commercial or financial outcomes of tasks come second to the well-being outcomes of individuals when tasks are being automated. This workplace is hard to imagine, and even harder to see today. Unfortunately many organisations are not at the stage where employee outcomes sit above financial or performance outcomes. Even though many workers today are faced with tasks which are both physically and mentally taxing and which, if automated, would likely alleviate many of the symptoms of stress we see today as a blight on our workforces. These tasks could in the future be fully automated to offer both the protection and security of healthy work to many professionals who today require considerable costly support just to get their jobs done.

The decision as and when to automate also requires ethical consideration when it requires potential retraining or workforce reduction. HR professionals in these circumstances are required to consider the broader issues at play than financial efficiency savings. In these scenarios a set of principles to guide decision making can help to improve outcomes, and also ensure that in delivering automation of tasks HR is considering all of its stakeholders beyond financial owner managers and investors. This includes delivering value for employees – too often left out of decision making on issues that directly impact them.

In the world of AI and algorithms there will be many decisions that HR practitioners need to make, and which will have long-term impacts on the quality and type of work undertaken in our organisations. Without HR we risk all decisions on automation in the workplace being driven by efficiency alone, and this cannot be a good thing for the future of the human workplace.