HR Tech: Useful Tools But Not For The Unwary  

Sue Turner OBE, Founder, AI Governance Limited and one of the world’s top 100 women in AI advises an informed approach to AI in talent acquisition.

I know it’s not December, but a recent e-newsletter got me thinking about the exciting opportunities that technology is going to offer HR leaders in 2024.

The newsletter from LinkedIn trailed “Recruiter 2024”, their amped up service which will include generative AI tools to make it easier to create hiring projects and surface potential candidates. Many other HR service and software providers are bolting on tech solutions that they say will generate efficiencies, reduce repetitive and mundane tasks and free up your team members for higher value work. Whilst these tools are useful, they bring problems too. So how can we get the best from technology whilst avoiding mistakes that could be career-limiting?

One example from Hilton Hotels shows the complexity. They have over 400,000 team members globally so efficiency in talent acquisition is important as they seek to source people who fit well in their culture of great customer service whilst also controlling costs. Over recent years they’ve tried AI-powered tools like HireVue’s video interview screening and Pymetrics neuroscience-based games. HireVue ran into trouble in 2020 when it was revealed that its algorithm, which was supposed to assign certain traits and qualities to job applicants’ facial expressions, was based on flawed science. HireVue changed its offering and continues to dominate the video interview market, although many candidates don’t like feeling judged by technology.

Question Your Tech

For HR leaders, there’s a clear lesson that it’s vital to ask probing questions of third-party suppliers when you are considering procuring AI tools. You need to understand:

• how the algorithms work
• what features are they using to make their prediction
• what data has the model been trained on and how does that relate to your data
• how has bias been mitigated
• how accurate will the predictions be in your particular domain
• how you will explain the results to candidates and employees.

You will need to be transparent about how you are using AI. In the UK and EU, the General Data Protection Regulations give people the right to opt out of automated decision-making, so you must let them know what you are using AI for, and give them an alternative route if they decide not to use the AI-driven part of the process.

One way to have more insight into the AI tools you’re using is to develop them in-house, which is what tech titan IBM has done. They proactively use their Watson AI platform to source applicants who match key success profiles. Using AI to predict who could be a good candidate, by finding patterns hidden to human recruiters, has enabled IBM to recruit more inclusively and increase diversity across its talent pool.

IBMers then extended their AI use to personalise Employee Value Propositions (EVPs), using data on the individual employee’s career path, preferences and performance. The result is tailored training programmes, career development plans and benefits packages. Rather than humans facing the daunting task of trying to create individual plans, the HR team members focus their attention on the higher value tasks of understanding the technology and the insights it produces. Similarly, Accenture’s AI-driven MyLearning platform recommends training and development opportunities for employees, increasing their engagement and skills development.

If you’re not a tech titan, developing your own tools can seem like a daunting task so take specialist advice so you can plan for successfully building your data science and AI team.

Authentic EVPs

We know that EVPs need to be authentic so the way you engage in AI ethics and governance should fuse your organisation’s purpose and values (including taking into account wellbeing and your diversity, equity and inclusion agenda).

We’re all working hard to improve diversity and equity in our hiring, but as we increasingly rely on tech tools and platforms, we need to be aware that we may be excluding people who are not on these platforms or who struggle to use the tools. For example, many assume that all young people will be equally capable of taking part in a video interview or online skills or personality testing. I know from my own experience of being CEO of a grant-making foundation, however, that not all young people are digital natives. Some struggle to use technology through language, reading or physical difficulties. Others simply don’t have the money to have always-on data so they spend a lot less time on mobile phones and other devices than we might imagine.

Use Without Expertise

I predict that 2024 will be the year that many HR leaders start using more AI tools without realising it, which will mean that some make mistakes through lack of expertise. You wouldn’t get in a car and start driving without putting your seatbelt on and having some understanding of the rules of the road, so don’t fall into the trap of using the exciting new tools from LinkedIn, or anyone else, until you’ve looked under the hood and satisfied yourself that you know what you’re doing.