Artificial Intelligence is a hot-button topic right now. In every industry, people are talking about how AI might impact their jobs. Will it make life easier? Harder? Will AI make them redundant?
In fact, my colleague Katt Silver and I recently spoke at a local SHRM chapter meeting about this very topic. Human Resources professionals are having the same conversations, but with one big difference. We have been using AI for a long time now—especially in recruitment and talent management.
Applicant Tracking Systems (ATS), which support the recruitment process for most organizations, already have some sophisticated Artificial Intelligence features, such as the ability to interpret a resume and send personalized messages to candidates. HR professionals have been using these AI features for years. And we’ve learned that AI has some advantages… as well as some drawbacks.
What is AI?
First of all, let’s define what we mean by AI. The official definition from the US Department of State is:
A machine-based system that can, for a given set of human-defined objectives, make predictions, recommendations, or decisions influencing real or virtual environments.
In practice, Artificial Intelligence is something of an umbrella term, covering a lot of different technologies. Some common forms of AI include:
- Process automation: Systems that learn how to perform complex process. Currently, most process automation involves a lot of manual programming by human operators.
- Machine learning: A system that can analyze huge volumes of data, identify patterns, and improve its own performance without being explicitly programmed.
- Predictive analytics: Similar to machine learning, this involves finding patterns in vast volumes of data. However, predictive analytics tries to work out probabilities of future events using data from the past.
- Generative AI: Systems that can create new content based on a user request. This includes tools like ChatGPT (which generates text) and Dall-E (which can generate images).
- Chatbots: An AI-powered front-end tool that allows people to engage in conversation and receive real-time responses; this can be used for customer service, answering basic employee questions, or interacting with job applicants.
All of these tools can be helpful in HR, especially in the area of recruitment.
How is AI being used in recruiting?
All of these systems described above can help improve recruiting. For example:
- Process automation: ATSs can automate recruitment workflow tasks, including candidate communications and interview scheduling.
- Machine learning: ATSs are trained on a huge corpus of resumes. This allows the systems to “learn” how to identify great candidates and produce more accurate results.
- Predictive analytics: Past data can help an ATS make predictions, such as whether a candidate is likely to remain in a position for 12 months after hiring.
- Generative AI: Generative AI can write letters and emails to notify candidates about next steps. It can also help hiring managers write effective job descriptions and interview questions.
- Chatbots: A chatbot front-end can allow candidates to ask questions about the recruitment process and manage things like interview scheduling.
Real-world examples of AI in recruiting
Our colleague, Tom Daly, at NFP wrote a great article on the Case Studies of AI in HR: A Glimpse into the Future of Human Resources. I wanted to highlight a couple of examples he included:
Unilever is a multinational company that receives thousands of applications each day. They faced two major challenges: processing applications in a timely manner and reducing bias within their hiring process.
In 2019, Unilever turned to AI to deal with both issues. They developed an AI tool that could analyze video interviews for entry-level positions. The AI used machine learning and predictive analytics to identify candidates most likely to succeed if hired.
Unilever credit AI with saving their recruitment team over 100,000 hours per year. Meanwhile, Unilever has won several awards for diverse hiring, including the 2019 Best Places to Work for Disability Inclusion.
Nestle is a well-known and one of the largest food and beverage companies in the world. They faced a major challenge in trying to hire large volumes of employees for their various global locations, which was a huge lift on their talent acquisition teams, in particular the task of scheduling interviews.
Nestle brought in conversational AI to help automate candidate screening and interview scheduling tasks. As a result, they saw a 600% increase in interviews and automated nearly 8,000 hours of recruiting work, answering 1.5 M candidate questions in the process, and scheduling 25,000 interviews.
However, not all such efforts end in success. In 2018, Amazon implemented a similar AI system with a goal of hiring new engineering talent. Amazon’s AI was trained on the library of all resumes received in the past.
There was one problem with this approach—engineering applicants were overwhelmingly male. Amazon’s AI was trained to believe that all engineers were men, so it automatically rejected female applicants. The system was soon shut down and Amazon apologized to affected candidates.
What are the potential risks and challenges of implementing AI in recruitment?
As we see from the examples above, Artificial Intelligence can be extremely helpful, but it’s not perfect. If you’re making a decision about using AI tools, you need to be aware of the pros and cons.
Pros
- Faster hiring: AI can process thousands of applications in mere seconds. This means a better candidate experience and much shorter time-to-hire.
- Better analytics: Automated systems tend to generate more data than manual processes. This means better management data (such as DEI metrics) plus more training data available for machine learning.
- Reduced workload for recruiters: AI can handle the time-consuming parts of recruitment, such as analyzing resumes and scheduling interviews. This frees up recruiters and hiring managers to focus on more important strategic tasks.
- Improved candidate experience: The shorter hiring cycle means that candidates don’t spend weeks waiting for a decision. AI also helps improve communication, which leads to a better overall candidate experience.
- Constant learning and improvement: AI tools are getting better every day. Over time, you can train systems on your internal data, which will produce even better outcomes.
Cons
- AI errors: AI is not perfect. Generative AI tools are prone to errors known as “hallucinations”, which can be hard to detect. Remember that all AI tools require close human oversight.
- Lack of human touch: The human element is an essential part of recruitment, as it gives candidates a taste of your organizational culture and values. If recruitment is fully automated, you miss the chance to connect personally with candidates and develop a foundational relationship.
- Risk of bias: As Amazon discovered, AI tools can actually increase bias in your hiring process. It’s important to have good oversight and regular quality checks.
- Possible compliance issues: Individual states such as California, Colorado, and New York are enacting legislation for the safe use of AI tools. These laws may create an additional compliance burden for your HR team.
- Increased costs (for now): Researchers have found AI-driven processes are typically more expensive than hiring a human to perform the same role. This may change over time, but it’s important to remember that “automated” does not always mean “cheaper”.
Conclusion
Artificial intelligence has been part of recruiting for a while, and the next generation of AI tools will hopefully make hiring even easier by freeing up your teams from administrative and repetitive tasks to focus more on strategic initiatives.
That said, you will always need skilled recruiters and hiring managers on your team; AI should be used as an enhancement to your team, not a replacement.
Want to learn more about AI? Check out these other articles: