Ethical Use Of AI In Recruiting
Debunking the top myths about AI in Hiring
Posted on 08-20-2024, Read Time: 6 Min
Share:
Highlights
- Ethical AI can enhance hiring by providing data that aids human decision-making.
- Incorporating AI into your evaluation process offers a more objective analysis of the skills and competencies that best predict success.
- Creating clearer regulations for AI and hiring will take some of the pressure off companies that have to try to make sense of multiple standards when vetting potential partners and vendors.
AI in hiring is a big topic with a lot of opinions, but when 75% of workers admit their negative perception of AI in hiring is due to a lack of understanding, it’s safe to say education on the topic is needed.
Hiring AI is not a mystery box open to interpretation, it’s rooted in decades of scientific research and data. There is a science behind ethical recruiting AI.
How AI is Being Used
Ethical AI can enhance hiring by providing data that aids human decision-making. Available AI solutions include automation, conversational AI, and assessments, varying in complexity and impact. Deployments like chatbots streamline candidate interactions, while complex solutions like AI-driven assessments, integrated with an ATS, evaluate skills and fit for roles, offering significant benefits.The Benefits of AI in Hiring
- Supports decisions with data
- Creates a fairer candidate experience
- Supports DEI initiatives
- Unlocks candidate potential by assessing skills and competencies for roles
Debunking the Top Myths and Best Practices for AI in Hiring
● Myth: AI decides who gets the job.
AI solutions, including assessments, provide important decision support for hiring teams. Algorithms should be built to support human decision-making in a structured and consistent way that mitigates human bias. These solutions empower smarter decision-making — and should never be the decision maker.
● Myth: AI replaces a face-to-face interview.
AI-assisted recorded interviews actually get you to the live interview faster. On-demand interviews typically replace the resume review or phone screen phase — speeding the process along.
● Myth: AI is changing in uncontrollable ways, so you can’t trust the results.
While AI algorithms can certainly be learning “online,” that is not the standard in hiring assessments. Assessment models should be trained and tested in the “lab”, then locked for deployment. It wouldn’t make sense to use a different algorithm day-to-day when evaluating candidates for a role. And, with the spreading use of Generative AI, it’s worth noting that these algorithms should be deterministic, meaning the same input should produce the same output every time.
AI solutions, including assessments, provide important decision support for hiring teams. Algorithms should be built to support human decision-making in a structured and consistent way that mitigates human bias. These solutions empower smarter decision-making — and should never be the decision maker.
● Myth: AI replaces a face-to-face interview.
AI-assisted recorded interviews actually get you to the live interview faster. On-demand interviews typically replace the resume review or phone screen phase — speeding the process along.
● Myth: AI is changing in uncontrollable ways, so you can’t trust the results.
While AI algorithms can certainly be learning “online,” that is not the standard in hiring assessments. Assessment models should be trained and tested in the “lab”, then locked for deployment. It wouldn’t make sense to use a different algorithm day-to-day when evaluating candidates for a role. And, with the spreading use of Generative AI, it’s worth noting that these algorithms should be deterministic, meaning the same input should produce the same output every time.
The Value of Bias Mitigation
Hiring teams have a responsibility to create candidate experiences that treat and evaluate all candidates equally — relying on data-driven evaluation instead of subjective resume reviews.A resume review does little to prove a candidate’s value as an employee, and the process invites unconscious bias to creep in. Bias isn’t just discriminatory behavior towards those with certain names, alma maters, or age. It can also present as familiarity bias — an affinity for people from the same hometown, who cheer for the same team, or even have similar interests.
Incorporating AI into your evaluation process offers a more objective analysis of the skills and competencies that best predict success. For example, an assessment built specifically for a product manager can evaluate candidates on the skills that make successful employees for that specific role.
Data and algorithms should be audited to check for bias and mitigated accordingly. In addition to third-party audits, companies should rigorously test every algorithm before it’s put into production to minimize bias. Once live with customers, they should be checked regularly to monitor for bias and assist with diversity hiring efforts.
Hiring and employment are already well-regulated, but creating clearer regulations at the intersection of AI and hiring will take some of the pressure off companies that have to try to make sense of multiple standards when vetting potential partners and vendors. Computers allow us to formalize fairness/bias requirements that we define in an automatic way.
Rather than simply incentivizing an algorithm to predict an outcome (in this case, a job-related competency), we simultaneously penalize it for having significant demographic group differences in the outcome.
Author Bio
 |
Dr. Lindsey Zuloaga is the Chief Data Scientist at HireVue, managing a team that builds and validates machine learning algorithms to predict job-related outcomes. As an academic researcher with a Ph.D. in Applied Physics, she has performed novel experiments and data analysis, resulting in scientific publications with applications in medicine, sensing, and signal processing. |
Error: No such template "/CustomCode/topleader/category"!
