Home » Knowledge Hub » HR AI » Love or hate? AI in HR.
Love or hate? AI in HR.
26 April 2023 HR AI
Story by
Sue Turner OBE Founder and Director, AI Governance Limited
The recent flurry of headlines about Artificial Intelligence (AI) has raised its profile dramatically. Whether you are proficient in using AI or a novice suffering from FOMO, it’s important to go beyond the hype and recognise that, while AI has the potential to revolutionise HR, it also raises significant concerns.
Fundamentally all artificial intelligence tools do two things: they find patterns hidden in large amounts of data and then use those patterns to predict, personalise or automate. They are already being used in all aspects of HR, from automating repetitive hiring tasks like CV screening to providing performance insights for line managers and personalising training to improve employees’ experiences.
Figure 1 – AI tools are being used in all aspects of HR
Why we might start to love AI
A survey of 259 HR leaders at US companies with 100+ employees found most were already using AI. The majority planned to increase its use in the year ahead, not surprising as it can help us increase efficiency and cut costs.
Unilever, for example, uses AI to screen job candidates for its graduate programme, with results showing time and cost savings whilst the quality of its hires has improved.
I have supported a client in the logistics sector to use AI to make the workplace safer. Using computer vision they can detect where employees are not following the correct safety procedures so supervisors can take action to improve safe ways of working. This level of monitoring would never have been possible without technology.
Companies are increasingly using sentiment analysis, a type of AI, to produce real-time reports on how positive or negative employees are feeling towards the business. One large business has stopped using the blunt instrument of an annual staff survey as they now have 24/7 insights from analysis of their internal communications channels.
This sort of data, coupled with other performance data, can be used to personalise the employee experience by creating bespoke learning plans, providing real-time feedback and connecting employees with the resources they need. For line managers these insights are invaluable, enabling them to focus their attention on employees most at risk of leaving or in need of support.
But on the other hand…
Fear of AI taking our jobs is frequently cited by people at all levels in HR teams and it’s a genuine threat. It’s likely that many HR roles will change dramatically over the next five years and there will be job losses. I would argue that’s a reason to be in the vanguard of using these tools rather than hoping they will not affect you or your team.
Without knowing how to use AI well, and with very little regulation for now, mistakes are being made which can impact on businesses and people’s lives. In 2022, three former Estée Lauder employees received out of court settlements after they claimed they were not adequately informed about how HireVue’s computer vision technology was used to select them for redundancy.
We’ve also seen Amazon publicly embarrassed by trying to use AI to make hiring decisions only to find that the data used to build the model was biased.
Our ability to collect and analyse vast amounts of data about employees can lead to difficult ethical decisions. We can track employees and predict their future behaviour through monitoring how they spend their time, what they do on social media and even analysing their friends and family. But should we? Or does this cross the line between employee privacy and surveillance?
The London School of Economics found AI was better at hiring than humans as it was faster, increased the fill-rate and recommended candidates with a greater likelihood of being hired. Job candidates, however, were overwhelmingly negative towards AI hiring and they viewed organisations using it as less attractive than those hiring through humans.
Another study tested AI’s ability to analyse employees’ work and generate feedback. AI identified more errors and made more recommendations to correct mistakes and improve skills compared to human managers. Employees who received AI-generated feedback achieved 13% higher job performance but only when they thought the feedback had come from a human. When employees were told the feedback came from an AI tool they achieved 5% lower job performance.
Conclusion
It’s no surprise that many HR leaders are unsure whether, or when, to ramp up their use of AI tools. But AI is getting into all aspects of our work so whether we love it or hate it, we need to learn to live with it.
Often the best combination is for AI tools to provide insights and for humans to provide the interface with job candidates or employees. HR leaders have the opportunity to lead this change so my advice is:
- educate yourself on what your data really means
- find out how AI works and how to harness its power in your organisation
- combine the best aspects of technology with uniquely human skills.
Sue Turner is one of our panelists on the AI webinar which is on Tuesday 9th May commencing at 11am. Click here to register.
Also take part in our AI/HR survey – click here – it’s only three minutes of your time!