“Artificial Intelligence breaks technology hiring process”

By Scott Hamilton

I experienced something quite unique this week while talking with a friend and prior co-worker. I will keep his name, the name of the company he works for and the recruiter’s name anonymous, because the story I am about to share is frankly embarrassing. They have been recruiting people for highly technical roles for more than a decade and have always had great success, up until the beginning of the month. In early August they interviewed what they thought were four different individuals for software developer roles at two different companies and hired all four individuals on an annual contract.

I know it seems boring so far, but trust me the story gets quite interesting. You see these four people were actually one person, utilizing an AI and probably stolen identities to obtain work. The person that was hired for all four positions had generated four separate AI personalities, with different hobbies, ages, races and personalities. He utilized the AIs during the interview process for all four positions through the same recruiter, who participated in all the interviews. He was able to pass technical interviews by four separate teams of six or more engineers and came across as an experienced software engineer.

In one case he interviewed for two positions on the same team and interviewed twice with exactly the same set of engineers. They had no idea they were interviewing the same person, because he was relying entirely on AI generated responses for all the interview questions, including personal questions about hobbies and interests. The engineering team was impressed with his communication skills and technical knowledge. He had four impressive resumes and impressed every interview team. He was hired for all four positions within a week of applying and started work for both companies around the second week of August. Each position having a salary range of $120,000 to $160,000.

He reported to work for the 100% remote positions and attended regular work meetings and completed all assigned work tasks for all four positions for around two weeks, before he made a critical mistake. He used the wrong AI generated video person on a meeting. In other words the wrong person showed up for a work meeting. The team would have never caught on had he used the same race for all his personalities, but a guy clearly from India showed up in the place of an Asian. Needless to say this triggered an investigation into the entire hiring process. It was then that it was discovered that he was holding down at least four jobs through the same recruiting agency and possibly more. In all likelihood had he kept better track of which AI was working which job he would probably still be collecting a pay-check. You see the AI systems have gotten so good at technical communication and software development that with the proper guidance from a good AI engineer they can easily replace even senior level software developers.

This raises a couple of questions. The first is since he was meeting the needs of the company and fulfilling the duties of the job role, by unconventional means, should he really be fired from the role? The second is how can technology companies insure that they are hiring real people and not AI systems in a completely remote work environment. I know after hearing about this that our company will change the interview process to insist on face-to-face interviews, unless the person we are interviewing is known personally by someone on the team.

There is actually another way that AI is impacting technology hiring and it stems from the fact that a majority of technology jobs now require the use a AI and the companies providing AI services are paying extremely high salaries for the right developer skills. This places extra pressure on smaller companies when attempting to hire technology staff as it is difficult to compete with $300,000-500,000 annual salaries offered by companies like Google for top tier engineers. Keep in mind that as AI become more advanced it gets harder to determine if you are talking to a real person or a computer, and you might just wind up doing business with an AI. Until next week stay safe and learn something new.

Scott Hamilton is an Expert in Emerging Technologies at ATOS and can be reached with questions and comments via email to shamilton@techshepherd.org or through his website at https://www.techshepherd.org.