Greenhouse CEO Daniel Chait recently issued a blunt warning: hiring is stuck in an "AI doom loop." Candidates use AI to game the system — resumes, portfolios, take-homes. Employers respond by adding more automation. The loop feeds itself. Nobody wins.
In a recent survey of 2,200 candidates, Greenhouse found 67% use AI to improve their chances, 45% embellish resumes, and nearly 1 in 4 fabricate work samples. It's a race to out-fake each other, where the winners aren't always the most qualified — just the most polished.
The solution isn't less AI. It's better use of AI.
As the founder of Vouch — yes, an AI-native hiring system — I agree with Chait. We saw similar findings in a recent, smaller survey of candidates. But I'd reframe the conclusion.
Why the loop exists
The doom loop exists because too many teams use AI as a shield designed to deflect volume, instead of a lens to understand people more clearly.
When resume parsing is treated as a filter, and keyword scoring replaces context — it invites gaming. Candidates adapt. Signal degrades. The problem isn't AI. It's how narrowly we've used it. What's missing is context, transparency, and better judgment.
That's why Vouch was built around a different principle: AI should assist all the way, but never replace human judgment.
We don't screen out people because a bot doesn't find a match. We use AI to define what you're hiring for more clearly, surface relevant context on candidates faster, and automate logistics that slow teams down. The human steps — interviews, synthesis, final judgment — are preserved, even enhanced.
- When you replace resume filters with contextual review, signal improves.
- When you make sourcing smarter, not noisier, outreach conversion jumps.
- When you pair AI insight with human conversations, confidence in decisions actually goes up.
The real irony
Most legacy hiring tools were built to enforce process, not empower judgment. But great hiring isn't about checking boxes. It's about spotting patterns, surfacing potential, and acting with clarity. That's where AI should help — not get in the way.
So yes, the doom loop is real. But the answer isn't retreat. It's reset.
What candidates are actually doing
Our internal research shows 80% of job seekers now use AI in their search. The top use is keyword optimisation and tailoring materials to beat the filters — not to express themselves.
"I use AI to make sure my CV hits the keywords. It's the only way to get past the first filter.
"
It's not that candidates prefer automation. It's that the process demands it. When the system favours polish over signal, people adapt. But it doesn't mean they feel seen.
That's the opportunity: to use AI differently. Not to filter faster, but to listen better.
How to break the loop without breaking trust
- Use AI for context, not conclusions. AI shouldn't decide who gets through. It should help you understand who's worth a closer look — ask for clarifications, guide you through.
- Anchor decisions in real signal. Not just job titles and credentials, but structured conversations, working sessions, and written tasks that show how someone thinks.
- Flip your filters. Don't just weed out. Design your process to surface in. Make space for depth, for edge cases, for non-obvious strength.
That's the kind of system we're building at Vouch to guide you through it.