Back to Blogs

ChatGPT: HR saviour or nightmare?

  • Publish Date: Posted about 1 year ago
  • Author: Handle Recruitment
AI generated image of people in office sat at computers typing

The above image was created using generative AI. The prompt was ‘People in office sat at computers typing - magenta, lime green and grey.’

It’s nearly impossible to look at the news without seeing a new development or think piece around AI. And recently, the focus has fallen strongly on generative AI.

If that doesn’t ring a bell, maybe you’ve heard of ChatGPT, or even Google Bard? These are open source (public) language tools which use machine learning to generate human-like responses to the inputs people give them. Much has been made of their many uses (both current and potential), but these technologies are complex and rapidly developing, so it can be tricky to know where to start.

To cut through some of this complexity, I sat down with Gethin Nadin, Chief Innovation Officer at workplace tech company Benefex, for a wide ranging chat about how generative AI tools can be used in HR and recruitment - and what to watch out for. 

Hi Gethin! Firstly, since you talk to so many people about this - what are some of the most interesting uses of generative AI in HR? 

Gethin: A lot of the work these technologies can do is lift admin away from HR, and give more space for making strategic decisions. Last year, HR professionals told us that they still spend about 30% of their time firefighting and responding to queries. The power of AI is that companies can collate their data together and potentially have these systems do much of the admin and answering work, on a personalised basis, leaving more room for the interesting work.

What are your thoughts on the more creative or strategic uses of generative AI, like in recruitment or content writing?

Gethin: AI and recruitment have a pretty murky background because ultimately data points have human fingerprints on them. For example, if your recruitment process historically has been biased towards men, ultimately when you put all that data in, you’re going to teach the machine to be biased towards men. So in my opinion with recruitment, things are still pretty murky.

The content area is interesting. Someone who’s not a very good content writer can write something pretty decent with the help of a programme like ChatGPT. Realistically a good content writer is still going to be better - but ChatGPT can bring up the levels of a poor content writer. It will upskill the unskilled. 

For HR and potential uses of generative AI, are there data privacy concerns - especially when concerning really sensitive data e.g. CVs or company financial data? 

Gethin: When we talk about companies using these technologies, companies need to create closed instances, and basically secure the parameters of their own data.

ChatGPT is open source - anyone can use it - you can go in there and ask who I am and it will tell you. You can also give it my address and it will learn that. Some people at the moment are using ChatGPT to write sales reports, marketing plans etc. but by doing so you’re giving it that data. You should never use something public like that for this kind of work - make sure your parameters are closed. 

Some people are concerned that generative AI is coming for our jobs. Should we be worried?

Gethin: We’ve always lost jobs to tech developments, but on the whole people aren’t very good at predicting how that will happen. If you look back in time to the invention of the car or telephone, what people were saying is exactly the same rhetoric as now. If you think about App development for smartphones, it's a huge business now - but it didn’t even exist 20 years ago. So I think people should be focused on the jobs AI technologies can create, rather than just being afraid of the ways technologies may change or disrupt current jobs.

I speak to a lot of HR people, and I get the sense that lots are picking up on fear, and missing out on opportunity. There are very different types of AI; self-driving cars are a big risk to human life, but people using generative technologies to create content for businesses aren’t. Overall, we need to learn to have more nuanced conversations about these things.


This article is included in this week's People Experience newsletter.
For more of this content and to receive your fortnightly summary, sign up below 👇

SIGN UP: The People Experience Newsletter