You are here

Will AI in Recruitment be the norm by 2023?

8 Dec 2020
Ewan Anderson

A study by in 2018 published a stat that suggested that 46% of HR professionals believed they would be using AI to a high degree by 2023. That's not all that far away, and I wonder if many of you felt your HR team were on track to be making the most of everything that AI can offer?

A recent survey run we ran at Eden Scott found that 90% of respondents did not use AI in the last year as part of their HR role. 

So where are we with AI really? What are people's confidence levels when it comes to trusting AI and the decisions it makes?

It certainly seems there is some level of scepticism amongst the HR functions across the UK when it comes to the adoption of AI.

Why, according to research from the Oracle/Future Workplace report, is it that workers in the UK are amongst the least excited by the prospect of AI compared to those from China or India for instance? 

What is it that is holding us all back? 

The prominence of artificial intelligence's in the rest of our lives is growing, so what's preventing those working in the HR profession?

Well, from our recent survey, it would seem the main reason is budget and a lack of leadership buy-in. 54% of respondents highlighted budget, and 16% indicated a lack of leadership buy-in as the top two barriers to technological integration.    

Photo by Possessed Photography on Unsplash


Trust is a crucial issue. If you don't trust the system, then it will be tough to convince your leadership team to take the plunge. 

Making decisions that will impact on peoples jobs and their future careers based on the data from a system is a leap of faith for many. A fact that seems especially true when you consider the issues faced by Amazon a few years ago and the gender biases their AI recruitment tool produced. 

Trust develops over time with a legacy of performance and positive results. However, the success of any AI system is the result of a programme of continual learning. AI requires good quality data to establish a hypothesis, a series of instructions and then an ongoing series of situations that can be rationalised to improve performance, continually. 

Data on its own won't establish bias. An AI programme responds to the conditions set at the start, and any prejudice that emanates from the data produced will reflect either past problems or issues with the initial conditions set.  

For instance, on further investigation, it is believed the failings in the Amazon system were a little more than ten years of hiring data that indicated a pattern of recruiting men.  

The problem seems to have been around the objective setting. They were trying to establish a recruitment system that removed virtually all the human touchpoints, leaving the decision making up to AI.  

But ten years of data from successful hires did not give full weighting to the cultural changes in terms of gender and ethnicity that have adapted hiring practices for the better. Their objectives seem to have tried to fast track the hiring practices based on skills and characteristics that, perhaps, no longer fit the mould. 

While they will have undoubtedly factored in gender scoring, the conditions set were not ranking the gender balance high enough to counter the systems overarching aim, which was to hire the best person based on the perceived conditions. 

While the system will continually learn and develop, setting the objectives and initial parameters are critical, as is having the HR team involved with the design of the programme at the very start.   


Photo by Alexander Sinn on Unsplash


The cost of a new system will always be an issue, especially as we now grapple with the economic fall-out from COVID-19. 

However, any initial outlay, which should include technical support as part of the HR team, must be considered against the potential benefits in terms of productivity and a more engaged team. 

The is no need to start with a complex AI hiring system, similar to that of Amazon, which seemed to jump several stages to try and automate the full hiring process. 

AI is making everyday tasks far more effective, supporting HR teams to deliver a more engaged service. Tasks like virtual onboarding, room bookings (when we finally get back in the office) and auto-filling registration forms are all tasks that take up time and impact on new employee experience. 

Similarly, people's learning and development journey can be enhanced by the learning of the many others that have gone before. Adopting a data-first mindset, according to Elizabeth Greene, Director of Global Learning and Development at ON Semiconductors, will make the whole experience for your employees more meaningful.  

AI can support by personalising the employee's learning, taking into account future development plans, job descriptions and experience of others in the business. 

Chatbots can also replace the need for HR teams to have to respond to every enquiry, whether that is training and development, onboarding or other HR enquiries. 

So while the cost of a number of these systems can be high, the return in time and productivity from your HR team will more than likely offset the initial outlay.  

The reality is, AI is becoming an essential part of our working lives, and it will support several HR functions in the future. However, with so many vital decisions to be made, the objectives need to be clear from the start, and it is clear that there needs to be HR representation on the development team.  


Ewan Anderson
Share this article
Advanced Search