EY: Collaboration between humans and AI in HR
In the last decade, the rise of artificial intelligence (AI) has been exponential. By the end of 2020, the AI industry is expected to have a total revenue of , with of emerging technologies having AI foundations as we enter into 2021. While the International Data Corporation (IDC) expects investment in AI to be impacted by the outbreak of COVID-19, the corporation still predicts that the worldwide revenue will surpass $300bn by , with a compound annual growth rate (CAGR) of 17.1%.
With this in mind, Business Chief spoke to Ray Joyce, Global Talent Predictive AI Lead, HR Services and , Global Talent AI Lead, HR Services at EY to discuss the rise of human and AI collaboration in HR operations.
Working in the industry, Joyce identifies two major trends that are currently occurring in the industry, “I think one major trend that is being shown at the present time is to provide HR and, in turn, organisations with greater insight to their data.” He highlights that harnessing machine learning at an organisational level helps identify trends in data which may not have been easily identified before. “From an employee perspective, the biggest trend is most definitely chatbots – which not only provide a fast and reliable place for an individual to have questions answered but also a safe environment, they can ask any question they like and get a personalised answer back in real time.”
While agreeing with Joyce, Gill does predict that the future will move away from using AI solely for insights in data and chatbots, and become more centred around automation. “This is already being seen in chatbots which are beginning to replace (in a limited fashion) Employee Self-Service portals and beginning to carry out transactions, such as a change of address. This level of automation will only increase as companies become more confident in allowing AI to complete tasks on behalf on an individual,” comments Gill, who adds that “for instance, a recruiter could raise a new requisition; as soon as this has been completed, the system has already reviewed the current talent pool(s) available and made recommendations on who would be a good fit – and more critically why. Process mining can review current HR processes and make recommendations on savings based on budgets, time and people.”
identifying that talent is the most expansive resource that an organisation has, and typically the most expensive. Joyce explains that, while the digital footprint that employees generate on a daily basis is significant, many companies do not always ‘know’ their people. AI allows talent teams to analyse the data in ways that they have never been able to in the past. With this insight also comes the ability to automate many tasks based on the data in a way that is not only consistent but also objective.” However, Gill comments that while this is certainly a benefit for organisations, “HR data is not always ‘clean’ - as any company that has done a data migration will know - so there may be tasks needed to ready the data. Once this has taken place, it will be of utmost importance that the rules and mechanisms used follow the organisation's inclusion, equality and diversity principles, so that trust can be achieved.”
While there are a lot of AI solutions out there, Joyce states that organisations must “see through the ‘sales’ and ensure that the solutions are tied to key metrics within the business case. Start small - but think big!”
Agreeing with Joyce, Gill explains that when it comes to developing an effective AI adoption strategy, it is important to “ensure that there is an ethics and bias review team in place which contains members of the HR team. Ensure that every step of the journey they are part of the team and understand the AI. Ensure HR knows that they own the solution and that it is their AI. In addition, do not plan or structure your strategy so that you have to have everything 100% accurate; AI needs to learn and to make mistakes, bring your employee base with you, let them have a say in what they want so that solutions are meaningful to them. It’s all fine having the most techy, whizzy solution, but if it is not what the employee or business wants, then basically it just turns into an expense and the budget could have been used elsewhere.”
Further to Gill’s explanation of an effective strategy, Joyce adds that “it is important that HR teams feel involved in the AI solutions. AI should not be ‘done to’ but ‘with’. Most of the data or processes that AI will change or replace will be owned by the HR or People teams and it is very important that the ownership stays with them. Letting the HR team have a say in how the AI will operate will help drive the adoption and take away uncertainty. It is important that content owners trust the AI to make the right decision or answer the question correctly, especially in an area like HR.”
Ultimately Joyce concludes that when it comes to the collaboration of humans and AI in HR operations, “An open mind is critical. Technology is changing rapidly and what may be true today can change tomorrow.” As a result, organisations need to be agile in their thinking. Joyce also highlights that emerging technology is not perfect to begin with, and there may be failures along the way. “Having strong stakeholders, a resilient culture, and accepting potential failures as a learning curve instead of a reason to stop will allow companies to keep moving forward, and then ultimately reap huge rewards.”