AI and HR and its potential pitfalls
As with all sectors, technology and the digital are changing the way HR processes are enacted. Managing employee records, communicating with employees and recruitment have all benefitted by utilising digital practices. This transfer has also been exacerbated by the pandemic, with more of the hiring and onboarding process necessarily moving online as well. There are clear efficiency benefits to these moves from a business perspective, but what are the implications of newer technologies that are currently maturing? How can AI be integrated into the HR workflow? What are some of the potential pitfalls? Let’s start with the basics.
The basics AI is a hot topic across sectors, but away from the flashy stuff like self-driving cars and winning eSports championships, what are the more practical applications from an HR perspective? In fact, you may already be using AI to recruit. Predictive tools candidate sourcing tools, such as those found on LinkedIn, make use of AI to make these kinds of decisions. It will usually do this by matching the requirements of the job to candidate CVs. This increases efficiency by saving recruiters from trawling through countless CVs by prioritising those most suited to the role. It’s important to remember that AI is there to recommend, not to make decisions. While the technology is advancing at an astonishing rate, it can only be as good as the data it is fed. When that data is something as subjective as a CV, there’s only so much it can do to rank candidates.
Its limitations Some have made the point that removing the human element from CV trawls allows for less bias in hiring, removing the human from an initial stage and therefore eliminating bias. However, AI can only work from the data it has, and if the data itself exhibits a bias then the AI itself will be biased. For example, in Florida, an AI was briefly used to determine the likelihood of a criminal to re-offend. The AI predicted a Black teenager, who was being charged with petty theft for trying to ride someone’s bike while running late to pick her sister up from school, as a high risk to re-offend, whereas a White 41-year old with a lengthy criminal record being charged with shoplifting tools was labelled low risk. The exact opposite ended up being true, with the Black teen having no new charges while the White man is now serving an 8-year prison sentence. This is an extreme example because the stakes are so high when employing technology like this in the justice system. However, this serves as a healthy reminder that no final decision should be taken by an AI and a reminder that it does have its limitations.
The future However, there is another level to AI which may come to bear more fruit over the next decade. If a company begins to collect data about things such as employee performance on various categories of tasks, AI could be used to assess internal candidate quality more effectively by assessing candidate competence at various aspects of a role. Of course, this would incur costs to implement, raises concerns about employee privacy and final decisions should still be made by humans, but this is the kind of thing that AI is developing into. It’s worth staying abreast of developments in this area, as well as ethical research, so when the time comes you can implement It into your HR processes effectively. Whilst we may not all be ready to implement a full AI alternative as our HR solution, we can make the most of technology that incorporates smart technology and boosts efficiency such as Breath. Learn more about how Breathe HR can transform your current HR solution.