top of page

AI in HR - What does the future hold, dancing Robots or powerful tools?

During last week I managed to go to the AI everything conference at the World Trade Centre in Dubai. My mission was to look out for current and future use of AI within the field of HR. Coming away from the event I have established some initial thoughts and feelings with AI's integration into the corporate world, and more specifically, the areas of HR it is likely to service. It did give me some food for thought regarding the following questions in HR:

What areas of HR could it advance?

From my walk around at the AI everything conference I saw a what the current approach focuses in the areas of HR:

1. HR automated Administration services

There were a couple of stands that demonstrated systems that were able to carry out HR administration services. By using chatbot technology as the interface, either through an actual robot or could be on an employee's machine, it provides the employee a 24/7 place for request and services. These include leave requests, letter requests, answers to policy and procedures, and on labour law.

What was interesting to note on some providers was the ability to conduct interviews and reviews via the chatbot. Programmed to ask questions based on Job descriptions, but also potentially with competency matrix as well. In this demonstration the focus was on internal recruiting, for transfers. But it demonstrates the continuation of the AI recruitment trend and could also lead into talent management area with conducting review processes. One example could be that the chatbot could lead the questioning for 360 processes.

From my research I understand that AI has been used to support the preliminary recruitment processes. This has been used in a proactive way in searching of candidates over online platforms, contacting them via chatbots in preliminary interview type questions to be able to rank in suitability. Also, it has been in a reactive way in recruitment by analysing submitted CVs, using chatbots in structured interviews, or face recognition techniques on video posted interviews with the result of cutting 1000s of applicants down to about 200 faster and more efficient than HR personnel. It is because the algorithm is ranking the candidates in order of suitability for further interviews that the recruitment team and line managers can process in a more traditional approach.

2. Face Recognition Technology

From my walk around there were approximately ten providers that would focusing on the use of facial recognition technology for the measuring employee engagement, and attendance. By scattering cameras around the office, it was demonstrating the ability of live monitoring of employees, tracking things like current moods, and movement. It is claimed that through the analytics the level of engagement can be ascertained by cross referencing this information with other employee data.

Knowing in theory this sort of information is no doubt useful, however, I imagine this level of monitoring raises several concerns for individuals, companies, and country privacy laws. It will be an interesting debate to be had and how that will play out in the next few years. I understand that there are such positions like Chief Ethical and Humane Use Officer are coming into business to look at these sorts of issues to keep the balance between the need of information against the levels of privacy.

3. Predictive AI

The last thing I noticed at the conference was the use of AI in predicting behaviours. This was seen within the area of Finance, looking at customer behaviours in relation to the bank’s products. Based upon the customers standard information (age, gender, family status, current salary, etc), and previous banking history the model could make a predictive percentage of how likely, or how receptive they would be to certain financial products (Loans, Credit Cards, Mortgagee). The next step would be to ascertain the likelihood of being able to keep to term of the repayment plan, their risk assessment.

This made me think regarding the area of performance management and the recent article I have read where IBM have managed to make attrition predictions up to 95% accuracy and understand when and why employees are likely to leave the company. With the relevant data points in your HR system identified, it could be possible to achieve several HR requirements, including talent pooling, succession planning, gap analysis for learning and development amongst others.

What about the HR data?

We need to be aware of the predictive nature of AI and that we are not at a "Hello Dave" moment, where the AI is a free-thinking mainframe capable of making the strategy decisions, we are a little way off that yet. I would be more incline to say that the 'A' in 'AI' should stand for algorithm rather than artificial. As to be truly artificial the system would learn not just within its own parameters but also via outside information and adapt and change the parameters.

What this means is it all still comes back down to data validity, and reliability, and how strong it is in your system. The AI is basically an algorithm-based program that establishes and learns patterns by the continuous input of data and monitoring between current and previous data. This will still be dependent on how well as a company you have kept your data is how applicable the AI output will be. It is still the case of 'S**t in, s**t out.'.

Another point is whether the system you currently have or looking to have contains the required data points for Data Scientists (yes that is what they are called) to use in establishing these programmes. Less relevant data points and data you have in the system the less valid the AI's output will be. It would be like trying to predict if an employee is going to leave your business and this programme did not included any form of engagement data points or data, say from engagement survey or company reviews.

Is the HR professional redundant?

The short answer is a no. However, as time goes on and as the integration of AI becomes prevalent inside HR departments there will be a shift in what the human part of a Human Resources team will need to be doing. It will open the way for HR to become true business partners, not being bogged down in the administration side of things. They will need to become more data analytical, understanding the patterns that the AI is spitting out to them, acting more as a validator to the data and discussing with Management in what they mean.

Also, there will need to be a branch of HR that can work with the Data Scientists so they can build more affective programmes by identifying key data points within their own company systems, and data points from outside sources. There are several positions being created such as Head of Business Behaviour, Chief Ethical and Humane Use Officer, and A.I. Trainer. There will need to be more as data science and HR become more blended, meaning the future for HR personnel will be needing to learn the understanding of programming, the limits of the current algorithms, and to correctly interpret the results the AI produces.

Final thoughts

In the end, AI will be coming into business, "resistance is futile", to coin a well know Star Trek phrase. However, at this moment, there are some single remarkable products and achievements out there in what I would consider the first phase in the mass market of its use, I have yet to see one complete system that is remarkable of utilising all areas of HR.

The software provider that will be king, will be the one, that in the short term, can integrate AI easily, affectively, and economically with current systems. In the long term it will be the ability to produce an all-encompassing Human Resources Artificial Intelligence System (HRAIS) that has a stylish and an easy user experience and is economical to install.

30 views0 comments


Post: Blog2_Post
bottom of page