Young people today are well-versed in the complexity and tedium of job applications, but despite the general sense that the process has significantly intensified, it is difficult to comprehend how rapidly recruitment is being automated. The cost of living crisis has made it necessary for many students to look for part-time work, but landing an entry-level role has become nearly impossible. The introduction and implementation of artificial intelligence (AI) in recruitment is not only transforming the inner workings of companies and human resources (HR) departments but also how jobseekers are treating the application process, and how it is affecting them in turn. Young people without established careers, who are attempting to enter the workforce for the first time, are discouraged by the unreasonable number of steps required to apply to any job.
To employers, automation is a natural process towards increasing efficiency: the more comprehensive the initial screening process is, the less companies have to invest, and the more time is freed up by employees to focus on other tasks. Nowadays, most job applications consist of different steps designed to whittle down the number of candidates as quickly as possible: CVs are scanned and different types of tests are used to measure a person’s skills, abilities and personality. Asynchronous interviews are also common, where a machine analyses a candidate’s recorded answers. While the screening process is inherently machinic and dehumanised, companies will often, but unsuccessfully, attempt to model it on human interaction. Instead of taking an interest in would-be employees, employers choose to use human-like AI bots, which makes the experience of applying to a job more like stepping into the uncanny valley. Only if a candidate is lucky enough to fulfil all the different criteria that the software used requires will they eventually get to speak to a real person.
Central to the problem of AI in job recruitment is that candidates are kept in the dark about those criteria and have no way of knowing when or how it is used to judge their applications. While the HR departments that employ AI tools are unlikely to know exactly how they work, candidates know even less, which exacerbates the frustration of applying to jobs. At all times, they must operate on the assumption that they are judged by a machine, and adapt accordingly. One such adaptation has been for job applicants to use AI tools to assist with all aspects of job applications, from highlighting the key words to use on a CV to answering interview questions. Ironically, companies discourage job applicants from using the same tools that they use to judge them. Instead, they must spend considerable time learning how to maximise their chances of success, which is a tedious business in itself that is made worse by the ever-changing nature of applications. While companies argue that increased automation improves efficiency, it raises the question of who is the true beneficiary. Rather than eliminating the human workload, companies are offloading it to job applicants, who have to tackle applications that are only becoming longer and more complicated. Beyond the time it takes to complete them, automation and the introduction of AI into job recruiting have more deeply psychological effects. Job applications, by their nature, necessitate that candidates put something of themselves into them, but this only has a dehumanising effect when machines value, measure and ultimately reduce their identities to a numerical score.
Whereas a cover letter was previously the most personal aspect of a job application, new technologies have introduced additional steps that encourage applicants to “just be themselves”. At the same time, increasingly complex and sophisticated machines judge them, which poses an inherent contradiction. Anyone who has had to complete these additional tests also knows how degrading they are, and as tools are constantly changing and developing, it feels like trying to play a game without access to a rule book. Even more humiliating is the fact that many of the jobs that employ additional tests to narrow down the number of candidates are entry-level retail roles that pay minimum wage. Due to the impersonal nature of job applications, applicants often resort to quantity over quality, which naturally leads to an increased number of rejections, each of which feels like a mental battering. Additionally, because AI is seen as a fairer and more unbiased alternative to a human recruiter (which is not necessarily the case), it is impossible for applicants to understand why they are rejected. While applicants are forced to tailor their applications to the job, the job will rarely be able to tailor the feedback to the applicant.
Learning how to apply to jobs has become a job in itself. To fit the criteria created by a machine, we have to remove all aspects of ourselves and our humanity from our applications, all while being told to act as humanly as possible. We are forced to feign authenticity while the process of recruitment becomes ever more automated. This is not a new phenomenon, however, as applying for jobs has always necessitated some form of inauthenticity. The reason that an AI bot can do the job of a human being is that the desirable traits in workers have been codified, with all human elements already removed. Inauthenticity is thus a necessary element of applying to jobs and promoting yourself as a potential employee more generally. Fundamentally, this points to the fact that workers are replaceable. We see this more tangibly in precarious working conditions that force workers to always be on the lookout for new jobs. It is not surprising that few young people are excited about entering the job market. Recruitment has become a one-way street, where would-be employees are treated by machines, like machines. Having to complete job applications can be tedious and frustrating, but having to constantly perform is exhausting. The acceleration of implementing new technologies in job recruitment has made this more dehumanising, as we are now performing to machines, not other humans.
