What candidates need to know
about AI in Hiring
The Top of the Stack AI Primer
Our AI Primer clarifies what candidates can expect from the AI systems now screening most job applications. Rather than trying to beat algorithms or trick systems, your emphasis should be on avoiding falling into the algorithms’ blind spots.
BUSTING MYTHS None of the mainstream Applicant Tracking Systems (ATS) have a feature that detects "this resume was written by AI”. Not Workday, Greenhouse, Lever, Ashby, or Oracle. Using AI to help draft your resume
will not get you immediately flagged in the system.
Am I being assessed by AI when I apply to jobs?
You should assume AI touches every application. From there, the question is how much human judgment sits on top of that AI use. Our research has found that about 80% of Fortune 500 companies put applications through active AI ranking.
The Main Players: HireVue & Pymetrics integrate directly into the ATS platforms used by their customers.
HireVue
HireVue is used by over 700 companies globally, including 60% of the Fortune 100, 30% of Fortune’s Most Admired companies, and 8 of the 10 largest U.S. federal agencies (as of May 2023). HireVue integrates with Workday, SAP, Oracle, Greenhouse, and more.
Pymetrics
Pymetrics clients include Goldman Sachs, JPMorgan Chase, Unilever, Accenture, BCG, Bain, EY, PwC, Blackstone, McDonald's, Tesla, and LinkedIn. Pymetrics integrates with Workday, Greenhouse, and Ashby.
ATS-by-ATS: What We’re Seeing & How It Should
Influence Your Search
Applicant tracking systems differ in how deeply AI is involved in the screening process and how much human review follows it. Here’s how six of the most common platforms work, and what each one means for how you prepare your application.
Greenhouse
Greenhouse has actively marketed its resistance to algorithmic overreach. A strong, structured, evidence-based, and personable application goes a long way with Greenhouse, since human review is more likely to truly be a part of the early stages of the process.
Lever
Lever increasingly screens candidates via a short structured exercise at the point of application (not just a resume parse), so you must take that initial screen seriously; it may carry as much weight as your resume itself.
Ashby
Ashby's AI is explicitly assistive, not decisive. A well-written, criteria-matching application has a good shot at surfacing here, since there's a real recruiter involved in every decision.
Note: Individual employers may also layer their own internal AI tools on top of whatever ATS they use. Some will disclose this on the application.
Workday
The legal theory in Mobley v. Workday centers on proxies like employment gaps, where the AI then makes assumptions about what those proxies mean. Employment gaps and being "out of the workforce" for a stretch may carry more algorithmic risk here than a human reader would apply. If you have a gap, address it briefly and proactively in the resume (and mirror that on LinkedIn) rather than leave it as a silent blank that a filter might penalize.
Bamboo
There’s lower algorithmic-screening risk here relative to other systems, but you’ll also less structured feedback and process, since it's not built as a dedicated recruiting engine. A clean, standard resume typically performs fine.
Workable
If you are asked to do a one-way/async video interview through Workable, treat it exactly like a live interview in terms of prep, as these get scored/summarized by AI before a human watches.