How to get hired by bots

AI has infiltrated every stage of the job search – but has it made recruiting better? Eliza Filby asked

Recruitment expert James Reed last week warned that “graduate jobs are no longer a given,” sparking talk of a “work apocalypse” and the decline of white-collar jobs. Reed saw openings drop from 180,000 after the pandemic to 50,000 this year, a surprising drop. His advice to young people in this white-collar recession? Find a career where you work with your hands, not just your head.

However, as always, context is important. How much of this decline is simply rebalancing after the post-lockdown hiring surge? How big is the impact on the increasing tax burden borne by entrepreneurs or the loss of confidence in the business world? Or the impact of AI? As Reed himself suggests, the answer is likely all of the above.

However, behind this debate lies a practical issue that companies and platforms like Indeed can actually control: how people get jobs. Talk to anyone looking for work and the biggest frustration isn’t the lack of roles, it’s the silence that often follows an application. Candidates apply to many jobs and never get a response, without knowing why. This can be very impersonal and demoralizing.

Consider how deeply AI has infiltrated every stage of the job search. Even the roles you see on LinkedIn or Indeed are powered by predictive analytics that assess whether you are “worthy” of the ad. Behind the scenes, the system secretly builds a behavioral profile to assess your suitability. Once you apply, your CV is scanned by AI for exact keyword matches, sensitive to your phrasing, structure and layout. You may then face an automated video interview where an algorithm analyzes your word choice, tone and pacing.

If you nail a human interview, the person across the table often receives an AI-generated document that summarizes your skills, estimates how long you’ll stick around, and assesses your “cultural fit.” The final decision may be in a person’s hands, but almost everything leading up to that decision is driven by an invisible algorithm that few people truly understand.

Those who know how to game AI, or at least work with it, will have the upper hand. On YouTube and Instagram, entire communities now exist to teach people how to outsmart bots.

One such creator is Sam DeMase, ZipRecruiter’s Career Expert, who has over half a million Instagram followers. He’s seen a sharp increase in the number of people looking to understand and “hack” AI-driven recruiting. The core message is that AI literacy is important.

Demonstrate human and technological fluency

He offers pragmatic, if slightly confusing, advice on how to make the algorithm work for you, from mirroring the language of job descriptions on your CV to adjusting the way you speak in automated interviews (no umms). But he warns there is a fine line. Be too formulaic and you will fail to stand out. The trick, he says, is balance: demonstrating both human and technical fluency. In the era of algorithmic hiring, being explicit about your IQ and EQ skills, with clear examples and evidence, is critical at all levels.

Tech companies have sold these systems to companies large and small as time savers, and they certainly are. But I remember a partner at a large firm proudly telling me that thousands of graduates applied to his firm every year. “But isn’t that a bad thing?” I suggest. “By attracting lots of people instead of the right candidates, you’re just creating more work for yourself.” It is this kind of inefficiency that technology companies are trying to solve. But perhaps the real solution is to make the problem smaller by actively deterring the wrong applicants. It starts with honesty. Too many companies polish their employer brand until it shines. An accounting firm that highlights charity volunteer days but hides the horrors of audit season is selling false promises and may be targeting the wrong candidates.

As Sam DeMase points out, the biggest problem is a lack of transparency. Why don’t roles list salaries? Why not explain the process, timeframe, and expectations? “Too often,” he said, “employers are unclear about what they want, resulting in confusion for everyone.”

We’ve made recruiting more efficient but far less humane. Until companies realize that recruiting is about connections, not screening, we will continue to mistake automation for progress.

Eliza Filby’s book, Inhericracy, is out now