If you’ve spent any time researching Delta Air Lines’ flight attendant recruitment process, you’ve probably come across endless discussions about the airline’s On-Demand Video Interview (ODVI).
Applicants obsess over camera angles, eye contact, facial expressions and whether they smiled enough while recording their answers.
But according to information presented to candidates before they begin the assessment, many applicants may be worrying about the wrong things.
Delta openly tells applicants that artificial intelligence can be used to evaluate interview responses and generate a score that is presented to recruiters as part of the hiring process.
AI Creates A Score Based On Your Answers
In information provided to candidates before they start the assessment, Delta explains:
“Artificial intelligence and/or machine learning creates a score or recommended score by checking if the content of your responses relates to the competencies and behaviors shown to be important for success in the role.”
In other words, the system is looking at whether a candidate’s answers demonstrate the skills and behaviors that Delta believes are important for flight attendants.
That might sound alarming to some applicants, but it’s not entirely different from how Applicant Tracking Systems (ATS) have been screening resumes for years.
The difference is that ATS software evaluates written applications, while Delta’s system evaluates interview responses.
Recruiters Still Make The Final Decision
Importantly, Delta stresses that the technology is not making hiring decisions on its own.
The airline explains:
“The hiring manager and/or recruiter can still make the final decision.”
Delta says AI simply provides additional information that can assist recruiters when reviewing applications.
That’s an important distinction because it means candidates are not being hired or rejected solely by an algorithm.
Instead, AI-generated scores become one factor among many that recruiters may consider.
The Biggest Surprise? The AI Isn’t Looking At Your Face
Perhaps the most interesting detail is what Delta says the technology does not evaluate.
For years, flight attendant applicants have worried that video interview software might be analyzing facial expressions, eye movements or tone of voice.
According to Delta, that’s not what’s happening.
The airline states:
“The artificial intelligence evaluates only the words used in your response.”
Delta goes on to say that the system does not analyze:
- Facial features
- Facial expressions
- Eye movements
- Tone of voice
Nor is the video used for facial recognition or identification purposes.
If that’s the case, then many applicants may have been focusing on the wrong aspects of the interview process.
What Is The AI Actually Looking For?
Delta comes up with this rather complicated explanation of how the system works, saying it replicates “the judgment of multiple expert human raters who evaluated thousands of responses to questions like the ones you are about to answer.”
The goal appears to be identifying responses that demonstrate competencies and behaviors associated with success in the role.
That means candidates may benefit more from providing clear, structured examples that showcase customer service skills, teamwork, and problem-solving abilities than worrying about whether their smile looks natural on camera.
The New Reality Of Airline Recruitment
The use of AI in recruitment is becoming increasingly common across many industries, and airlines are no exception.
With some flight attendant recruitment campaigns attracting tens of thousands of applicants, technology is increasingly being used to help recruiters manage the volume.
For aspiring flight attendants, Delta’s explanation offers a rare glimpse behind the curtain.
The airline has effectively confirmed what many applicants suspected all along: AI is helping evaluate video interview responses.
At the same time, Delta’s own guidance suggests that candidates may have been worrying about the wrong things. If the system is primarily evaluating the content of responses rather than facial expressions or eye contact, then what you say may matter far more than how perfectly you deliver it.
