Top 10 Things Every HR Practitioner Should Know About AI
This article is not a comprehensive guide to AI in HR. My goal is to simplify and condense my personal learning experience of artificial intelligence in HR into an article. I hope this article will help other HR practitioners interested in learning about AI expedite their journey.
So, without further ado, here is Part I of the top 10 things about AI that every HR practitioner should know:
1. What is AI?
Artificial Intelligence or AI refers to machines programmed to perform tasks and think like humans. It simulates human intelligence. AI is not a new concept, as it has been used in various domains for decades.
You might have already experienced AI in your everyday life in the following ways:
Robotics: remember the robotic process automation you thought about for your HR Service Center team? Yup, that is an application of AI
Speech recognition: that text-to-speech or speech-to-text functionality you’ve used on your phone? Also, AI
Image recognition: when you used to drop an image to Google image search to look for similar images? That is AI
Natural language processing: that survey tool you got a few years back that automatically grouped open comments in pulse surveys into themes? That was AI
Predictive analytics: the predictive attrition model your Analytics team produced a while ago? A simple version of AI
The point is that AI has been part of our lives for a long time. It is moving from our personal and consumer lives to our professional world.
2. What can AI do for HR?
There are differing opinions on what AI could mean for the HR field. AI has the potential to revolutionize HR functions and the way we work. Imagine if most of the repetitive and administrative tasks were taken off your plate this week. How much time would you save, and what other tasks could you focus on in your role? This is what AI can do for you and your HR team.
In HR’s daily context, picture having an assistant who has keenly observed your work over the past few months, noting every decision and choice pattern. Over time, this assistant can predict your reactions or actions in a particular scenario. In this case, this assistant is AI, which analyzes massive amounts of data to identify patterns.
3. How do I Adopt AI for HR?
AI is a powerful tool that can be leveraged by HR to solve business problems. However, it should not be adopted just for the sake of it. Like any other technology, it should be used only when it can deliver value and solve a specific problem.
To effectively use AI in HR, you must start with a clear problem statement and a well-defined objective. You should identify the business challenge you are trying to address, consider the constraints and parameters, and define the ideal outcome. Once you understand the problem, please work with your HRIS or IT team to determine whether an AI-based solution can help.
Remember, investing in AI without a clear business case and a clear return on investment is not a wise decision. So, always start with a business problem and an ROI case before considering the adoption of AI.
4. What does AI mean for HR?
It can be our best friend and our worst enemy. It can be a valuable ally by simplifying or eliminating the administrative and repetitive tasks we all face daily. However, it can also be a threat because it exposes formerly ‘confidential’ information that HR may have closely guarded. For example, job candidates can quickly create customized resumes that match our job descriptions using AI-powered tools. Social media is another source of new hacks on using AI tools to answer interview questions during live interviews. Additionally, compensation information that was previously considered confidential is now more accessible to our candidates and workforce.
The use of AI in HR is a game-changer. It alters how the HR department interacts with its stakeholders, modifies how we view our work, and redefines our value to the broader organization.
5. How can I get started with AI in my HR Team?
Most HR functions with their own HRIS and People Analytics resources are already making progress in adopting AI. The basics required for AI implementation are similar to what your HR Technology and People Analytics teams have done over the past few years. It simply comes down to the following key factors:
Accurate data, as AI relies heavily on data to function
Solid data privacy practices because all the data you use for AI needs to be compliant in how they are gathered and handled
Removal of biases: as AI is trained on your historical datasets, it can amplify existing and historical biases if not correctly managed.
Well-planned HR tech stack, as AI solutions need to integrate seamlessly into your current tech stack and be able to scale
6. Given the rapid advancement of AI in the last few years, should we wait and see what happens instead of adopting the technology now?
In the first part of this article, I expressed my opinion on the importance of not using AI just for the sake of it and instead adopting it only if it solves a business need. However, the opposite is also true. If your organization faces a challenge that AI can solve, waiting a long time is not a good idea. Inaction could be worse than adopting AI, a rapidly evolving solution, and improving it as it develops.
In the context of HR, an example would be automating specific processes in your Tier 1 service organization using AI, such as virtual agents that can answer common employee questions. This can create more capacity for your team and prevent the need to expend additional resources, leading to delays in finding a solution. Waiting at this stage could result in unnecessary resource expenditure and further delays.
7. How do I know if AI is the solution for my organization?
Very much like other technology solutions in the market, the best way to do this is to:
Define the problem you are looking to solve
Define what your ideal state is
Determine if AI can/should be a part of the solution.
Set parameters on the scope of the solution (e.g., which workforce population would be impacted by this?)
Create a pilot with 1-4 in mind and test out the AI solution
It is best practice for HR teams to only commit to technology solutions after ensuring they work in their environment and solve current and future challenges.
8. Will I lose my job to AI?
As an HR practitioner, it's essential to understand that your worth doesn't lie in the administrative tasks you perform or how quickly you complete them. The true value of HR lies in partnering with the business and supporting the workforce to achieve organizational goals. Manually approving timesheets every week only contributes a little towards achieving that.
While each situation is unique, most HR organizations won't see job replacements due to the advent of AI. That doesn't mean that HR jobs won't change due to the prevalence of AI in the workplace, but it's unlikely that machines will entirely take over anytime soon. Here are some reasons why.
HR departments have been under pressure to do more with fewer resources for years. However, with the help of AI, we can transform how we work by upskilling and reskilling employees currently focused on routine administrative tasks. This will enable us to provide our workforce with a more personalized and customized experience.
Every HR department has a long list of projects to complete and limited time and resources to accomplish them. AI has the potential to assist us by taking on more of the mundane and administrative tasks, freeing up our time to focus on projects that will shape the future of our workplaces.
AI is poised to help HR become more human. Often, HR practitioners find themselves caught up in the daily grind of mundane tasks, which can sometimes lead to a sense of detachment from the workforce we are meant to serve. However, imagine a world where you are not bogged down by paperwork or system work and can instead focus on the "Human" part of HR, which is our primary purpose.
9. Can I buy the AI technology and go from there?
Maybe. It’s not as simple as buying the first one you see. In addition to ensuring that the solution addresses your specific problem, here are some other factors you should consider when evaluating AI solutions:
What AI model is the solution built on? How was the model chosen and fine-tuned for HR?
How do you validate the performance of your AI models?
How is the AI model trained? Is it just on your company’s data or another dataset?
How does the model learn and adapt over time?
How does the AI model arrive at specific decisions or recommendations? Are there any “black box” components, or is the entire process transparent?
What data is needed for the AI model, and how is that data stored, retained, and handled?
How does the solution flag or address potential bias in your training dataset?
How does the solution gather data from your existing systems?
Do you use proprietary algorithms or rely on third-party libraries or tools? If third-party tools are used, how do you ensure their reliability and security?
If there's an issue or dispute regarding an AI-driven decision, how is it addressed?
This is not an exhaustive list—just something to get you started in the conversation.
10. Bonus: How can I use AI in my work today?
As you have reached this point on the list, I'd like to share some tips on how and where you can incorporate AI in your work, which can help simplify your life. I am sharing these tips based on my experience and not endorsing any solutions. Here are the tips:
If you have writer's block on emails, newsletters, or any copy text, ChatGPT can generate some ideas for you, and Grammarly also has an AI-based text improvement/writing functionality.
If you need to create a beautiful presentation, Canva is my go-to.
If you need images to be generated, Canva still takes the crown for me on this one; Blue Willow is also decent, depending on what you are looking to do.
If you need help analyzing a spreadsheet, you can try the ChatGPT Beta Data Analysis.
In a way, AI is like play dough; you need to explore and play around with it to find the best way to serve your productivity and solve your organization’s challenges.