Jobseekers may soon find themselves interviewing with AI instead of humans, as more companies turn to artificial intelligence to analyse candidates’ responses and generate follow-up questions in real time.
Since the pandemic shifted much of the recruitment process online, hiring tech has become increasingly sophisticated. Software that records a candidate responding to printed questions is fairly commonplace, but this only replicates a one-way conversation. Now, that same software can respond live using a synthetic voice, such as by asking follow-up questions or answering questions about the role.
In addition to simulating a more realistic, two-way conversation of a traditional interview, tools offered by the likes of HeyMilo AI and Ribbon can ask candidates about the desired skills and provide summarised feedback to hiring managers.
The idea is to make applicant screening more efficient by eliminating the need for staff to attend every interview, while also giving applicants the flexibility to choose interview times that suit them, which may be outside regular working hours.
Such software is equally beneficial beyond the initial screening stage. Arsham Ghahramani, CEO of Ribbon, told Bloomberg that 15% of interviews on his platform happen later in the recruitment process, such as during salary negotiation or when providing feedback to unsuccessful applicants. These are the more uncomfortable conversations that human staff may prefer to offload.
Job market in turmoil thanks to AI
It is not a particularly easy time to be looking for a new job, and while there are other factors, AI is part of the problem.
Job applications increased by 31% in the first half of 2024 compared to the same period a year earlier, and 45% of applicants are utilizing AI to assist with completing them. Naturally, using ChatGPT to answer the competency questions will speed up the process, enabling job hunters to apply for more roles.
This has become a burden for hiring managers, who have seen a spike in applications without a corresponding rise in qualified candidates. In the UK, each graduate job role sees an average of 140 applications, according to the Institute of Student Employers.
Recruiters are also becoming increasingly skeptical of AI-written applications, which often include generic phrases such as “leverage my skillset.” But the volume of applications continues to grow while the quality is declining, and now recruiters are turning to their own AI tools to sift through the CVs they receive.
At the same time, hiring teams are struggling with fake job applicants, who use deepfakes and generative AI to make their way through the hiring process. Their motives may include bypassing visa requirements, applying on behalf of a group, or conducting corporate espionage in preparation for a cyberattack.
Risks of using AI software in the hiring process
While AI hiring software, particularly the type that conducts real-time interviews, can be useful for recruiters, it also comes with risks. HireVue previously used visual analysis to evaluate candidates’ expressions and traits but rolled back the feature in 2021 due to concerns over its validity and bias.
Research has also found that many open-source AI recruitment tools are more likely to recommend men over women, especially for higher-paying jobs. Workday is facing a collective action lawsuit over allegations that its AI-driven applicant screening tools disproportionately reject applicants over the age of 40. Incidents like this may lead to increased regulatory scrutiny in the near future, with New York City already implementing a law requiring employers to conduct annual bias audits of any automated hiring tools.
Even setting aside concerns about bias, AI interviewing tools can malfunction. Several humorous TikTok videos show bots repeating themselves in loops, speaking gibberish, or interviewing each other.
Braden Dennia, an employer at the investment startup FinChat who uses AI tools for hiring, told Bloomberg that the software finds questions from the candidate “tricky to field.” He also acknowledged that some applicants drop out of the running once they know they will be interviewed by AI. According to Consumer Reports, the majority of Americans are uncomfortable with the idea of their interview answers being analysed and graded by an algorithm.