AI Cheating in Job Interviews: How to Spot It
Is AI helping candidates cheat? Learn how recruiters can spot AI-driven cheating in online interviews, how to fight back, and keep the hiring process fair.
The way companies hire people has changed a lot in recent years. More and more, job interviews are happening online instead of in person. This is because it can save time and money, and it allows companies to look for talent all over the world, especially if they are offering remote roles.
Along with this shift, artificial intelligence, or AI, has become a big part of how people find jobs and how companies find workers. People looking for jobs use AI to help them prepare for interviews and even during interviews.
While AI can be helpful in many ways, its use in job interviews also brings up some important questions about what is fair and what is not. If someone uses AI to pretend they know more than they really do during an interview, is that okay?
The move to online interviews has made it easier for candidates to use technology in potentially dishonest ways. In a face-to-face interview, it's much harder to secretly look up answers or get help from someone else. The digital environment of online interviews, however, makes these kinds of things more possible. Also, some people looking for jobs might feel that the hiring process isn't always fair.
As companies use more AI in their own hiring processes, candidates might also feel more justified in using AI tools to help them. They might see AI as a way to even the score or get an advantage in a tough job market.
There is a notion that AI is responsible for rejecting candidates. This misconception is often fueled by scam artists selling solutions on "how to beat the system," some career coaches, and even vendors marketing AI tools designed to cheat the process.
The people who create AI cheating tools often argue that they are empowering job seekers and helping to level the playing field, especially when candidates feel like they are competing against AI systems used by employers.
I’ve never quite understood the reasoning behind the justification, “If companies are using AI, then I’ll use AI to cheat.” This is especially puzzling when it’s based on the false notion that AI is actively rejecting candidates. I believe they’re simply trying to justify candidates cheating by using these arguments to ease any guilt or discomfort about it.
AI Tools Job Seekers Use to Cheat
One way job seekers are using AI is through tools that act like a teleprompter, giving them answers or suggestions in real time during an online interview. These AI helpers listen to the questions the interviewer asks and then quickly suggest what the candidate should say.
For example, one tool has a feature that does just this. It works by having the job seeker upload their resume, cover letter, and even the job description before the interview. This helps the AI understand what kind of job they are interviewing for and what their background is. Then, during the interview, the AI listens to the questions and pops up answers that sound like something the candidate would say.
Check out one example of cheating in this Instagram video.
These tools try to sound natural and use information from the candidate's own background to make the answers fit. Some people who make these tools say they are just trying to help candidates who might be nervous or not good at online interviews to show their true potential.
However, many people in the HR field think using these kinds of AI tools during a live job interview is not fair. They compare it to cheating on a test because the candidate isn't really coming up with the answers themselves. There's a big debate going on about whether this is really cheating or just using the technology that's available.
However, while AI-generated answers might sound good, they can sometimes seem a bit too perfect and lack the natural flow and personal touch of a real human response.
With the growing accessibility and user-friendliness of AI tools, it's likely that more job seekers will adopt them despite lingering ethical concerns. This trend could create challenges for companies in accurately identifying the most qualified candidates.
AI Assistance for Coding Tests
Another area where AI is being used to cheat is in coding tests, which are common in interviews for tech jobs. Companies in the tech industry have noticed that some candidates are using AI tools to help them solve coding problems during online interviews.
These tools can do things like write code, find mistakes in code, and make code run better. These AI systems have been trained on a lot of coding problems from websites like LeetCode, which means they are very good at coming up with solutions to these types of questions. To deal with this, some companies are even thinking about going back to having coding interviews in person so they can better see if a candidate is really doing the work themselves.
There's a lot of debate about whether using AI for coding tests is actually cheating. Some argue that it's just using the tools that are available, similar to how developers use AI assistants like GitHub Copilot in their daily work.
Amazon, for example, has told its recruiters that candidates can be disqualified if they are caught using unauthorized AI tools during interviews. This could even result in the candidate being banned for several years or for as long as the company keeps its records.
On the other hand, some companies are becoming more comfortable with candidates using AI and are even starting to test how good candidates are at using AI tools.
The question of whether using AI in this way is wrong is complicated because in the real world of software development, AI tools are becoming more and more common.
Other Potential AI Cheating Methods
Besides AI teleprompters and coding assistants, job seekers might use AI in other ways to try and gain an unfair advantage. One method is to use voice modulation software. This type of software can change the sound of someone's voice, possibly making them sound more confident or hiding their accent during phone or video interviews. This could help a candidate make a better first impression without truly representing their natural speaking voice.
Many recruiters are now dealing with fake candidates, even with the use of AI filters. As AI continues to advance, it’s getting harder for recruiters to tell if they’re seeing a candidate’s genuine skills or ones enhanced by AI – or whether the candidates are even real in the first place.
In the future, we’ll see companies doing more thorough background checks to make sure they’re hiring people who are exactly who they say they are.
How Recruiters Can Spot Cheating
Recruiters have a few ways to spot cheating, but it’s not always easy to catch.
Observing Candidate Behavior During Interviews
Even with all the advanced AI tools available, there are still ways for recruiters to spot if a candidate might be cheating during an online interview. One thing to watch for is the candidate's eye movement. If their eyes are constantly looking to the side or down, it might mean they are reading from another screen or notes that are not visible to the interviewer. While it's normal to look away sometimes to think, a consistent pattern of looking off-screen could be a sign of cheating.
Another clue can be how long it takes a candidate to answer a question. If they take a very long pause before answering, especially if the answer is very well-spoken once they start, it could mean they are using an AI to generate the response.
Answers that sound too perfect, very generic, or don't include any personal stories or examples can also be a red flag. AI-generated answers might lack the specific details and personal touch that usually come from someone speaking from their own experience.
Technological Methods for Detection
Besides just watching and listening, recruiters can also use technology to help them spot cheating during online interviews. One simple method is to ask the candidate to share their entire computer screen. This allows the interviewer to see if they have any other applications or browser windows open that could be helping them with the interview, like an AI chatbot or a document with pre-written answers.
However, as demonstrated in the video below, some tools on the market have already addressed this issue.
Recording the interview (with the candidate's permission) can also be useful, as it allows the recruiter to go back and review the video later if they had any suspicions during the interview.
There is also special software called AI-powered proctoring software that companies can use. This type of software can do things like check the candidate's ID to make sure they are who they say they are, record their screen and webcam, and even use AI to watch for unusual behavior during the interview.
Some of these tools can even detect if the candidate is using AI to generate their answers. For instance, they transcribe their responses and use an AI-powered plagiarism checker to determine if the answers were generated by artificial intelligence.
For coding tests or written assignments, recruiters can use AI-based tools to check if the text or code looks like it was written by AI or if it's very similar to other things found online.
Is AI the new Excel? Shopify just made AI skills a baseline job requirement. This shift isn’t just coming—it’s here. Read this if you want to stay relevant: AI Skills Now Required: Is Your Job Safe?
Learn how to harness AI effectively and get better results every time. You can grab the book on Amazon (Paperback, Hardcover, Kindle) —no matter where you are!
Fighting Back AI Cheating (Cheapest to Most Expensive)
Low-Cost Strategies
1. Setting Clear Expectations and Policies:
The most basic and cheapest way to fight back against AI cheating is to be very clear with candidates about what is allowed and what is not during the interview. In the invitation and at the start of the interview, recruiters should state whether using AI tools or getting help from others is permitted.
Identified Insights: Simply telling candidates the rules can make some think twice about trying to cheat. It also gives the company a clear reason to disqualify someone if they are found to be using AI in a way that was not allowed.
2. Requiring Full Screen Sharing:
Another no-cost strategy is to ask the candidate to share their entire computer screen, not just the video interview window. Once they share their screen, recruiters can ask them to briefly show any open applications or tabs.
Identified Insights: By seeing the candidate's full digital workspace, the interviewer can potentially spot obvious signs of cheating, like an open chatbot window or a document with interview answers.
3. Paying Attention to Eye Movement and Delays:
Recruiters should be trained to watch for certain behaviors during the interview. For example, if a candidate's eyes are moving in a way that suggests they are reading from another screen, or if they take very long pauses before answering questions, these could be signs they are getting help. These observations don't cost anything but require the interviewer to be attentive.
Identified Insights: By being observant, interviewers can pick up on clues that might suggest a candidate is not answering truthfully or is getting outside assistance.
4. Asking Follow-Up Questions and Probing Understanding:
Instead of just accepting the first answer a candidate gives, recruiters should always ask follow-up questions. They should ask the candidate to explain their reasoning, give examples, and go into more detail about their answers. For technical questions, asking the candidate to explain their code step by step can reveal if they truly understand it or if they just got the answer from AI.
Identified Insights: This method tests the candidate's real knowledge and ability to think for themselves. AI can often give a good initial answer, but it might struggle when asked to elaborate or explain in detail.
5. Designing Questions That Require Personal Experience and Critical Thinking:
Recruiters can create interview questions that are harder for AI to answer. These could be questions about past experiences ("Tell me about a time when...") or questions that ask the candidate to think critically about a problem or situation. AI is not very good at making up believable personal stories or showing real critical thinking.
Identified Insights: By focusing on these types of questions, recruiters can get a better sense of the candidate's actual skills and how they approach problems, rather than just getting a general answer from AI.
6. Incorporating Live Coding or Problem-Solving Exercises (with Explanation):
For technical roles, a good low-cost strategy is to include a live coding or problem-solving exercise in the interview. Ask the candidate to write code or solve a problem in real time and explain what they are doing as they go. This makes it much harder for them to rely on AI without showing that they understand the process.
Identified Insights: This directly tests the candidate's skills and their ability to think and solve problems in the moment. It's a more active way to assess their capabilities than just asking theoretical questions.
There have been numerous instances of North Koreans (spies) attempting to enter the interview process, and one unconventional method has reportedly proven effective in identifying them. Simply ask the candidate to say something critical about Kim Jong Un. While this approach is certainly not the most professional, it is said to be in use by some and appears to yield results.
Mid-Range Strategies
1. Utilizing Features in Video Conferencing Platforms:
Many video conferencing tools have features that can be helpful in spotting cheating. For example, some platforms can show if the candidate's attention is focused on the interview window or if they are looking at other things. Recruiters should also make sure to record the interviews (with the candidate's permission) so they can go back and review them if they had any doubts about a candidate's honesty.
Identified Insights: Using the features that are already built into common video conferencing software can provide additional ways to monitor the interview without needing to buy new tools. Reviewing recordings can help catch subtle signs of cheating that might have been missed during the live interview.
2. Implementing Free or Low-Cost Plagiarism Checks (for written or code-based assessments):
If the hiring process includes any take-home assignments, written answers to questions, or coding tasks, recruiters can use free or low-cost online tools to check for plagiarism or AI-generated content. There are many websites that offer these services, and they can help identify if a candidate has copied their work from somewhere else or if it looks like it was written by an AI.
Identified Insights: These tools can help ensure that the work submitted by candidates is their own and reflects their actual abilities, without requiring a big investment in specialized software.
High-End Solutions
1. Online Interview Proctoring Software:
For a more robust solution, companies can invest in online interview proctoring software. This type of software often includes features like verifying the candidate's identity, recording their screen and webcam, locking down their browser to prevent them from opening other websites or applications, and using AI to monitor for suspicious behavior during the interview.
2. Onsite Interview For Every Role
Trust is critical at work, so companies might consider going back to in-person interviews for the final steps of hiring. While this can be expensive and take more time, it's the hardest way for candidates to cheat.
Another option is to hire people called invigilators who can watch the candidate in person while they do the online interview or assessment. This could happen at a neutral location like a library or co-working space. While it costs money to hire these people, it might be less expensive than bringing all candidates to the company's office.
Identified Insights: These alternative strategies can provide a higher level of security for certain situations, but they also come with their own costs and challenges.
Keeping Interviews Fair in the Age of AI
It's clear that AI is changing the world of job interviews, and this includes new ways for job seekers to try and cheat. Recruiters need to be aware of these methods and take steps to make sure the hiring process stays fair for everyone.
This means using a mix of strategies, from setting clear rules and carefully watching how candidates behave to using advanced technology like proctoring software. The goal is to hire people based on their real skills and potential.
As AI continues to get more advanced, recruiters will need to keep learning and adapting to make sure they can still accurately assess candidates in online interviews. The focus should be on finding the best talent in an honest and ethical way.
Exploring Remote Proctoring and Interviewing Software
For a more robust solution, companies can invest in online interview proctoring software. This type of software often includes features like verifying the candidate's identity, recording their screen and webcam, locking down their browser to prevent them from opening other websites or applications, and using AI to monitor for suspicious behavior during the interview.
Here's a table summarizing some of the proctoring software and their approximate costs (where available):