Right now, applicants are using AI to draft, tailor, and blast out applications, while employers are using AI and automation to screen, rank, and run early interviews. I understand why each side defends its choice, but the damage starts when everyone spends their energy trying to get a document through a filter, and nobody actually learns how the person does the work.
To me, a polished CV has become weak evidence of competence, just as awkward writing is weak evidence of the opposite. In November 2025, Greenhouse reported that 70 percent of U.S. hiring managers believe AI helps them make faster, better decisions with fewer recruiters. But confuses speed with quality: a hiring process can easily get quicker while the actual evidence of capability gets thinner.
When I speak with boards, I tell them the fix is practical. Instead of relying on automated filters, use a short, real-world test built around a decision the job actually requires: what would change the candidate’s mind, what they would verify for themselves, and what they would refuse to hand over to an AI tool. Give promising talent a small sandbox with one clear problem and a short deadline, and make sure a named leader owns the final hiring judgment.
And if you are early in your career, firing another hundred applications into the same automated funnel is a losing bet. Build something small enough to finish, and put it where a leader who actually understands the problem can see it. Real work provides evidence that a keyword screen isn’t built to collect.
The question I leave with boards is simple: if your front door is just two AI models optimizing a document, how will you ever recognize a builder who refuses to play that game?
I advise companies and leaders on navigating the changing job market. Contact me to arrange an on-site workshop for your board and leadership team.