Where AI Can And Can’t Help In Talent Management

Everyone’s trying to figure out how to bring artificial intelligence into different parts of their business operations. Eightfold AI research reveals that 92% of HR leaders say they’re moving forward with AI in at least one major area.

The only problem? AI isn’t perfect. In fact, AI can cause serious corporate snags. This is especially true when AI is brought into the talent management process, which encompasses sourcing, interviewing, hiring, onboarding, and more.

It can be tough to identify when and where you’re going to run into AI-related challenges when it comes to talent management. To help you determine where AI belongs in HR—and, perhaps more importantly, where it doesn’t—read through the following suggestions. They’ll help you see how you can apply AI to get the most benefit while limiting risk.

1. Use AI to streamline your new-hire training.

Bringing new talent on usually involves a significant amount of training. The good news is that you don’t have to create training manuals, video script outlines, worksheets, or other necessities alone. Generative AI products like ChatGPT may not be terrific at adding the human element, but they’re great at constructing outlines and gathering information. So, you’ll be able to leverage generative AI to complete your first drafts faster. Once you have those drafts in place, you can flesh them out as needed.

Writing for HR Daily Advisor, Lin Grensing-Pophal says that AI can be particularly helpful for smaller companies with limited training personnel and resources. She gives a good example of a small fitness business leaning into ChatGPT: “For example, someone in the fitness industry looking to build an employee handbook might start with ‘What information should be included in an employee handbook?’ Without having to read several blogs or books about handbook construction, the AI can give them a starting place they can build from.”

In other words, AI is better suited to be an informational training assistant rather than your own personal writer.

2. Steer clear of AI when recruiting.

Despite the fact that 79% of employers use AI or automation to recruit and pick top candidates, you probably want to stay away from this practice. Though it might seem like a clever, reliable way to pinpoint the best applicants, it’s not foolproof. On the contrary, AI is known for being biased since it is created by people. If you are not diligent in double-checking AI’s suggestions, you could make unintentionally discriminatory hiring decisions.

Dr. Dan Harrison, founder and CEO of Harrison Assessments, explains why it’s wisest to shift away from using AI for recruitment purposes. “When used for screening or hiring decisions, there is increased legal liability, questionable benefits over systems designed for assessment, and greater potential for a negative candidate experience and employee turnover,” he says. “For example, it could penalize women candidates because they are statistically more likely to leave the workforce to care for their children. Or AI could unintentionally have a bias against older people simply by penalizing people with too much experience.”

The point is that AI doesn’t have human discernment. The easiest way to stay out of hot water is to make sure you have human oversight for all your recruiting efforts. You may be able to automate and personalize some aspects of the candidate journey. However, you should reconsider allowing AI to source or evaluate job seekers.

3. Lean into AI to allow people to self-serve.

When you’re in HR, you’re pulled in a million different directions. AI can help fetch information or find answers for applicants and new hires so that you don’t have to be everywhere at once.

A good way to figure out how to delegate some of your tasks to AI is to make a list of all talent management duties. Be specific, and try to break each duty down to its smallest task. Once you’ve completed your list, you’ll start to see where AI could be an asset. Remember: AI should not be utilized for tasks that require empathy, compassion, or nuance. On the other hand, it’s exceptionally well-suited for grabbing information or making straightforward suggestions based on specific inputs.

Interestingly enough, 81% of people say they want more self-service options, but 40% of businesses think they have enough. Why not meet people’s expectations using AI?

4. Ensure humans handle layoffs and terminations.

AI should not be involved when you have to let people go. Whether you’re firing one person or expecting a mass layoff, don’t plan to involve AI in terminations. If you’re not sure who to fire (in the case of a major layoff event), avoid using AI software products alone. As mentioned before, AI can make the wrong decision and can have unintentional bias.

For instance, a company with discrimination issues may see a higher turnover rate for people of color. The Washington Post pointed out that AI could misinterpret this data. The algorithm might think people of color are more likely to resign anyway and recommend letting more people of color go.

Is it difficult to say goodbye to employees, especially under challenging circumstances like the sudden need to downsize your workforce? Certainly. Nevertheless, it’s best to deal with this issue personally and reserve AI for other jobs.

There’s quite a bit AI can do for talent management, but there are quite a few things AI shouldn’t touch. Knowing when to set AI in motion will help you minimize its risks and maximize its benefits.