Did your AI job ad add requirements you never agreed to hire for?
A small team needs someone to handle customer bookings. The AI draft sounds polished, but now it asks for a degree and three years with a particular software package. Who decided those were necessary? That is a hypothetical example, not a result I have tested. Greenhouse's existing documentation says its AI can draft a job post from a few details. A separate feature suggests interview scorecard attributes using the job title, description and existing attributes. An unexamined requirement in the description could therefore influence what the team asks AI to suggest next. Before generating the ad, I would have the hiring manager write down what the person must be able to do on arrival and what the team will teach. Compare the draft against that note. Delete additions nobody can justify, then check the interview scorecard against the same note rather than treating the polished ad as the authority. Greenhouse also says generated job posts are not automatically checked for bias. Human review still has a job here. Who on your team can remove a requirement before it becomes a reason to reject someone?
Comments
Put what you will teach in the published ad, too. Someone reading it after a shift cannot see the hiring manager’s internal note. If they can handle bookings but have never used your software, they may close the page before you ever hear from them. If it is genuinely the plan, say ‘We will teach you our booking system during paid training’ rather than leave them to guess whether ‘experience preferred’ means they have a chance. I would ask AI to make that distinction explicit, then have the manager confirm the promise. Removing a requirement from the interview scorecard cannot help someone who already decided not to apply.