Estimated Reading Time: 0
AI is often approached with hesitation, partly because some AI software, particularly tools using facial recognition technology, has perpetuated harmful biases in the past. And so many companies are reluctant to include a third-party AI platform to mediate hiring processes and risk potential liability.
But depending on individual company protocol, the interview process involves varying degrees of rigor and may demand several interviews over months – particularly in organizations vetting large numbers of applicants. Using AI to automate and streamline critical parts of the hiring process can generate significant time and cost savings.
But there are still various hurdles companies must address when adopting AI, including concerns about transparency, accountability and explainability, as well as uncertainty around potential for bias and the impact of emerging legislation.
The History of Artificial Intelligence
AI is often framed in sensationalist terms that remove it from its context. Unaware of its history, many people mistakenly see AI as a sudden departure from its technological predecessors.
In fact, AI research dates back to the 1950s and has been a subject of interest for decades. While recent advancements in machine learning, natural language processing, and other areas have led to significant progress, AI is not entirely new.
Understanding AI through the lens of its history offers a framework for demystifying AI, addressing the “fear of the unknown” by dispelling the myth that AI is wholly unprecedented and unpredictable.
Data processing, for example, is a feature of AI that precedes the invention of the computer, beginning with punch cards in 1750.
Of course, modern AI is far more advanced than the punch card system, but it often fulfills the same function: sorting and processing data.
Concerns about AI Transparency, Accountability & Explainability
AI is a hot topic today because of its potential to transform almost every aspect of our lives, from healthcare to education to technology. But AI transparency is a common concern related to AI and substantially impacts users’ trust in and willingness to use AI products.
AI transparency involves being open about data handling, the model's limitations, potential biases, and the context of its usage. AI models are often categorized as ‘black box’ or ‘white box’ models – ‘white box’ is transparent and explainable, meaning you can see clearly how a decision was made, while ‘black box’ AI is not.
In March 2022, researchers conducted a large-scale analysis of how black-and-white-box models performed on a broad array of nearly 100 representative datasets, spanning domains such as pricing, medical diagnosis, bankruptcy prediction, and purchasing behavior.
They found that for almost 70% of the datasets, the black box and white box models produced similarly accurate results. In other words, more often than not, there was no tradeoff between accuracy and explainability: A more-explainable model could be used without sacrificing accuracy.
“There’s no reason an organization should have to sacrifice transparency and explainability for accuracy,” explains Will Rose, Chief Technology Officer at Talent Select AI. “Making sure your AI vendor provides a white box solution is absolutely critical to long-term accountability – especially as we’re seeing more and more states consider AI legislation.”
There is also simply fear of the unknown to consider.
“The biggest misconception about AI in hiring is fear of the unknown,” Michael Campion, Ph.D., Herman C. Krannert, Distinguished Professor of Management, said in a recent interview with Talent Select AI on The Future of AI in Hiring. “The way these processes work is very simple and logical. They’re just tools for objectively scoring job-related information.”
Uncertainty around Bias & Emerging AI Legislation
In 2019, Illinois was the first state to impose legislation against AI interviewing programs measuring expression, tone, diction, and body language, limiting the parameters of their services, due to concerns of potential bias and adverse impact.
In 2020, Maryland followed suit, enacting a similar law prohibiting AI video interviews without express permission from candidates.
With recently-passed legislation in New York and Colorado – as well as additional AI legislation being considered in more than 17 states – it’s vital that organizations partner with AI vendors that can provide transparency and accountability to ensure absence of bias.
As public acceptance of AI grows and appropriate policy is enacted, use of AI throughout organizations is also expected to grow. According to 2023 research from Zippia, 65% of recruiters currently use AI, 67% report an improvement in the hiring process upon implementing AI, and 96% of senior HR professionals believe AI will “greatly enhance talent acquisition and retention.”
Additionally, 48% of managers openly admitted to having bias, and 68% of recruiters said they think AI will remove unconscious bias from the hiring process.
“AI provides an automated way to objectively and consistently score job related information, and it does so very effectively,” explains Dr. Campion. “It can actually be fairer than human judgements.”
The Talent Select AI Approach
Talent Select AI uses Natural Language Processing (NLP) to provide an unbiased psychometric assessment of each candidate – a measurement of their capabilities, personality traits, and core skills – all from the words they use in the job interview.
Building on decades of established Industrial Organizational (IO) Psychology, our explainable, white box AI platform has been heavily validated to ensure absence of bias.
“If an AI model helps make a hiring decision, it’s critical that we can understand and articulate why that decision or recommendation was made,” explains Rose. “If an AI vendor you’re considering can’t do that, you could be opening your organization up to serious liability.”
At Talent Select AI, we work with some of the world’s leading IO researchers to ensure the efficacy of our technology, and we are committed to providing transparency, accountability and explainability across our products.
Ready to see how Talent Select AI drives efficiency across the hiring process?
