The Evolving World of AI in HR

This week I had the pleasure of speaking at The Skyscanner Global HR Offsite on the topic of AI in HR. I opened my talk around the broader evolution of AI and specific terminology, at the bottom of this article you can find some videos on the technical detail, regulation or how the tech companies are evolving from an investment perspective from my go-to experts.

In terms of the evolution of AI in HR I see this in 3 separate areas:

1.    The traditional use cases such as anomaly detection, OCR screening of documents or scanning of receipts.

2.    This has progressed into the arena of AI making decisions, for example CV screening. This brings the ethical and regulatory challenges that the likes of Amazon and Google have famously faced in the talent space. My view here is that you still need a human to make the final decision.

3.    However, where it looks like AI is today and will be for some time is the role of an assistant or co-pilot. For anyone of you that have used ChatGPT it is the use case of being able to rapidly close the gap on the information you need, to do your job. What this will do for an organisation is allow their people to be more strategic and do less…non-strategic work!

76% of HR leaders believe that if their organization does not adopt and implement AI solutions, such as generative AI, in the next 12 to 24 months, they will be lagging in organizational success compared to those that do. Gartner

 How does an HR department maximise the value of this co-pilot functionality? It all comes down to context which again is where the experts in HR are critical. We constantly ask our clients how their people objectives link to their business objectives. It is often nestled in an annual report and many firms build this into their risk management however it was created by a human.

Whilst the value in AI is summarisation of information, generation of collateral and interaction with information and decisions, it is a huge waste of time without context from the HR expert themselves.

This reinforces the point that whilst AI will replace some tasks in a role, it requires strategic people to give the context to deliver the desired outcome. It is why I am huge advocate of everyone starting to experiment, even if it is co-pilot built in Bing on Microsoft Edge (almost all of you reading this will have access). Think about those dull, repetitive tasks that ultimately end in you making a data-driven decision, what are they? This is where your co-pilot can start assisting….

So, you know your people objectives, how can AI practically help today? 

There are some fantastic organisation out there and with a viewpoint that the foundation model layer will be commoditised, it is the application layer that will be receiving increased levels of VC backing. I wrote at the start of this year that 2024 is the year for the AI application after the rise of the foundational models in the last few years and even just in HR, this seems to be playing out.

The core HRM systems such as Workday and SuccessFactors will continue to bring AI into their platforms however it is the likes of Moveworks as a general HR Co-pilot or SkyHive speeding up interview processes or Eightfold.AI helping prepare your workforce for the future that will be interesting to track and see if they really bring measurable impact. At Aon we are taking the broader Human Capital and Risk Capital Co-pilot approach to help our clients to be better informed and advised in order to get to better decisions, at pace.

Looking a little into the future and particularly in these applications, they are no longer going to be static, but will change and adapt on the fly becoming polymorphic. So we will no longer need traditional campaign driven experiences in communications, these apps will simply rebuild themselves in ‘microsite’ format as they learn about you. Exciting stuff!

For those looking at this space I would recommend a few resources (they will probably be out of date in a few months!):

What are Large Language Models?

AI vs Machine Learning

Machine Learning vs Deep Learning

What is NLP?

20 Minute VC – What is the future of the AI Technology Market

Hogan Lovells The EU AI Act | What you need to know (includes broader insights into global regulation)

Please reach out with any questions or thoughts! david.kirk@aon.co.uk

Julia Clement Alastair Cockrem Salma de Graaff Jade Baker - thank you for having me!