In 79.4% of cases, artificial intelligence correctly identifies the most suitable candidates for advertised positions, according to a recent study conducted by DigAÍ in partnership with a Brazilian researcher from MIT.
The survey analyzed interviews conducted via WhatsApp and compared the scores assigned by the AI with the final decisions of the managers. The result was that, in 8 out of 10 cases, it classified as "above average" precisely those candidates who would later be approved in the selection process.
This precision reflects AI's ability to assess behavioral signals that often go unnoticed by human recruiters. According to Christian Pedrosa, founder and CEO of DigAÍ, the technology's goal is not to "catch" the candidate, but rather to translate reactions that, when analyzed together, offer a more complete and accurate reading of the professional.
"This type of analysis helps HR teams identify professionals with greater adaptability, consistency, and a predisposition to collaboration—key qualities, although difficult to capture in conventional processes," he says.
How does AI-powered recruitment work?
The methodology combines computational emotional intelligence, language analysis, and statistical models that identify behavioral patterns. In audio, for example, almost imperceptible vocal signals are observed, which are then cross-referenced with trained databases to recognize characteristics associated with professional performance.
In practice, this set of analyses allows DigAÍ to assess cultural alignment, clarity, and coherence of responses, even in situations where there is a contrast between the content said and how it is said. Overly rehearsed answers, a stiff tone, and an artificial posture, which have always been noticed by experienced recruiters, are now becoming even more evident to AI systems.
On the other hand, in companies, technology represents the chance to reduce biases, improve decision-making, and understand candidates more accurately, going beyond the so-called "gut feeling" during the interview.
“Technology expands what we can see. When we cross-reference what is said with behavioral patterns, we can understand the quality of reasoning, beyond the response, and how the candidate supports what they claim. It is an evolution that brings transparency and fairer decisions,” concludes Pedrosa.
