Could AI hiring filters reject qualified candidates before a human sees their CV? A Google DeepMind warning raises fresh questions about automated recruitment.
![]() |
| Google DeepMind’s reported warning shows why AI can speed up recruitment but still needs human oversight when screening job applicants. Image: JM |
JM Desk — August 12, 2026:
Could an AI hiring system reject a qualified candidate before a human recruiter ever sees the CV?
A reported warning from Google DeepMind has raised that question, highlighting the growing risks of automated recruitment.
The company’s AGI Safety and Alignment Team reportedly created a special form for applicants to help ensure their CVs reached a human reviewer.
An internal document reviewed by Bloomberg reportedly warned that there was a “non-trivial probability” a CV could be screened out incorrectly or take too long to reach the team.
Google disputes that its systems incorrectly filter applicants. The company says it aims to recruit the most qualified candidates and that there are no shortcuts to getting hired.
Still, the episode points to a wider problem with AI-powered recruitment.
AI tools can help employers process large numbers of applications, identify information in CVs and reduce the workload for recruiters.
But the same efficiency can become a weakness when an automated system makes an incorrect decision.
If a CV is filtered out before a recruiter sees it, there may be no opportunity for a human to reconsider the decision.
That makes human oversight especially important when AI is being used as the first gatekeeper in the hiring process.
There is also an emerging AI race between employers and job seekers.
Companies are using AI to screen applications, while candidates are increasingly using generative AI to tailor CVs and write application responses.
DeepMind is reportedly warning applicants against relying too heavily on AI-generated answers because such responses can sound repetitive and similar.
The result could be an unusual cycle in which AI-generated applications are assessed by automated systems before human recruiters enter the process.
That raises another question: how much of the hiring decision is actually being made by people?
The concerns extend beyond one company. AI recruitment systems have faced wider scrutiny over transparency, bias and potential discrimination.
Such concerns do not mean AI has no place in hiring. Used responsibly, it can help recruiters handle repetitive work and manage large volumes of applications.
The key issue is what happens when the technology gets it wrong.
For applicants, an automated mistake can mean losing a job opportunity without knowing why.
For employers, the lesson is straightforward: AI can assist recruitment, but human review remains essential when automated decisions can determine who gets a chance.
If applicants need a special route to make sure a human sees their CV, the recruitment industry may need to take a closer look at what “efficient hiring” really means.
