When AI Crosses the Line: The Risks of Using AI to Answer Video Interview Questions

As popularly said with the ease of everything comes its disadvantages as well, same goes with AI as well.

Today over 50% of applicants use AI to enhance their answers, people using AI for opportunities can be helpful but harmful as well, like using it for

Making your cv ATS friendly

Finding niche opportunities

For interview preparations, can be helpful and is ethical as well.

But using it during the online assessments or interviews is harmful and unfair, as there are AI tools to identify the use of unfair means as well which can not only

Restrict the availability to current opportunity

But also blacklist the candidate from company’s database as well

Closing the door to multiple opportunities

 

Moreover, the person sitting at the other end during the interview can also easily judge from the person’s

Facial expressions

Eye-movements

Can easily switch to behavioural questions or do cross questioning to verify that the answers are genuine or not

 

So this can be easily understood through an example:
“If asked what is sustainability, can be easily answered with the use of AI”
But if the question is switched to “what three steps would take if hired for this role to ensure that you contribute towards the company’s goal of achieving sustainable work environment,” then this is what a person can answer but AI cannot support it without proper context of the role and company’s working.

The unusual pauses or silent breaks before every response can also be an easy indication of person using an AI tool during the interview process.

Moreover, in most of the cases if an HR asks you something, the main motive is not to get a well-polished answer filled with jargons rather the point he is emphasizing is how you as an candidate are able to put forward your point/response in a way that can be easily understood and is relatable to the query or question of the person.

Sometimes a genuine statement “that I’m not familiar or well-versed about the topic currently, but will surely learn or research over it” can be a much better option rather an AI generated answer just to look that you know everything which is not even possible in the real-world scenario and the other person is also not expecting that as well.

The use of AI is not only limited to applicants or job seekers but also to the company/recruiters

So, it’s easier for them to detect the use of any unfair means as compared to the one using in the interview/assessment

Some popular tools include: HireVue, Talview, Mettl, Honorlock etc.

 

These tools help the recruiters in identification through

Screen monitoring

Voice tone

Eye movement and facial expression

Webcam monitoring and copy paste evaluation etc.

Certain online test/assessments works only while on full screen not even allowing to change screen during the assessment.

Moreover, the recruiters now don’t even need to monitor or filter candidates, the tool or platform used itself is capable enough to

Monitor tests/assessments

Blocking and rejecting an applicant in b/w of an interview if it senses the use of any unfair means.

However, some-tools or gadgets are there that are hard to monitor or detect, some

of the most commonly used include:

Alexa/Siri in-built feature that nowadays come with most the TWS (truly wireless buds), that automatically detects the questions without even requiring the person to provide any command to the AI tool/platform used

However, there’s a catch in that as well, just an easy human based step that’s you can ask the person to disconnect the Bluetooth device and use the device mic for some time or rest of the interview.

Tools like Eleven Labs & Voxbox are used for voice cloning during the answering which is one step ahead than the previous one. As it doesn’t even require the person to speak only lip-syncing is required in it.

But there’s a catch in it as well as the AI tools doesn’t fumble or think, the answers are too quick, correct and confident to be considered as genuine or true.

Moreover, tools like: Deep face, Snap Camera, ManyCam are used by applicants, where the person answering is using a mask of an applicant to answer the questions on his behalf.

Talking about the things to identify its usage include: no or very less blinking of eyes, too stiff emotions and reactions, irregular lip-syncing

So, if there are plenty of ways to use AI as an unfair means, then plenty of ways also exists to identify and tackle the misuse of these technologies or tools.

It depends on the recruiter/interviewer to be aware of such tools or technologies that can be used as unfair means

And the platforms that can be used not only to automate the process but also to make it error free and authentic by analysing and minimizing the use of any unfair means to as less as possible.

 

By Mohd Hamza Nadeem

Human Resources Team, SCORR Ltd.

Published on: 06th July 2025

0 Comments

Your email address will not be published. Required fields are marked *

This will close in 0 seconds