Google sells AI tools that can help HR teams analyse resumes and identify candidates. Now, one of its own DeepMind teams has reportedly warned applicants that relying on automated screening could prevent some applications from reaching a human.
The episode does not establish that Google’s hiring systems are broadly unreliable, as the company disputes that characterization. But it highlights a growing tension in recruitment: businesses are turning to AI to process candidates faster, even as concerns grow over how automated systems handle people whose experience does not fit neatly into machine-readable criteria.
Google DeepMind Team Flags Screening Risk
Google DeepMind’s AGI Safety and Alignment Team, which works on risks associated with advanced AI, reportedly created a special form for candidates applying to its open positions.
According to a document viewed by Bloomberg, the team told applicants that its application system had a “non-trivial probability” of screening a CV out incorrectly or taking too long to reach the team.
Completing the form would ensure that “a real human” on the team saw the application, according to the document. The instruction was reportedly accompanied by a request not to share the document widely.
Google DeepMind disputed the suggestion that its systems incorrectly filter applicants. A spokesperson said the special form was designed to bypass recruiter review and send resumes directly to the team, while stressing that there are no shortcuts to being hired.
That distinction matters. The reported warning is evidence of a concern raised by one DeepMind team, not proof that Google’s recruitment system routinely rejects qualified candidates incorrectly.
AI Is Entering Hiring
The broader shift is measurable. SHRM’s 2025 Talent Trends research found that 51% of organisations surveyed use AI to support recruiting. Among the uses reported, 44% involve screening resumes and 32% involve automating candidate searches. The most common recruiting application was creating job descriptions, at 66%.
The appeal is largely operational. Among HR professionals whose organisations use AI in recruiting, 89% said it saves time or increases efficiency. Another 36% said it helps reduce recruitment, interviewing or hiring costs, while 24% said it improved their ability to identify top candidates.
Google itself is part of this trend. Its Workspace products are marketed to HR teams with AI features that can analyse resumes, provide summaries to help identify qualified candidates and streamline recruitment workflows.
The business case, therefore, is not difficult to understand. When hundreds or thousands of applications arrive for a role, automated tools can help recruiters narrow the field before spending scarce human time on individual candidates.
Efficiency Meets Human Judgment
Google DeepMind’s own published recruitment process illustrates why the human element remains important. The company describes an initial recruiter conversation, two or three skills interviews, final interviews with team leads and leadership, followed by a hiring-team review against the criteria for the position.
The reported DeepMind episode sits somewhere between those two approaches. The special form did not eliminate Google’s recruitment process or guarantee an interview. It was intended to make sure an application reached the team for human consideration.
That points to a practical limitation of automated recruitment: the quality of the final decision depends partly on what happens before the human reviewer gets involved.
A candidate’s experience may be highly relevant but difficult to capture through predefined criteria. Career changes, unconventional backgrounds or transferable skills can create challenges for systems designed to identify particular signals. The evidence available here does not establish that this happened at Google, but it explains why the question of human review matters.
AI Hiring Bias Problem
There is also a broader governance issue. The International Labour Organization’s 2025 analysis of AI in human-resource management warned that systems can be undermined by unclear objectives, biased or incomplete data and opaque programming.
The organisation said these shortcomings can distort decisions, reinforce inequalities and expose employers to legal and ethical risks.
There are concrete examples of the stakes. The US Equal Employment Opportunity Commission said iTutorGroup’s hiring software automatically rejected more than 200 qualified applicants in the United States because of age. The company agreed to pay $365,000 to resolve the case. The episode concerned alleged age discrimination, rather than a technical screening error, but it demonstrates why automated employment systems can carry consequences beyond efficiency.
For employers, the lesson is not that AI should be removed from recruitment. The data shows why companies are adopting it. The more consequential question is where automation should stop and human judgment should begin.
Human Review Becomes Valuable
The DeepMind episode is revealing precisely because it does not prove that AI hiring is inherently defective. Instead, it shows how difficult it can be to balance speed with judgement.
AI can reduce the administrative burden of reviewing applications. But if organisations use it to determine who receives human attention, the screening stage itself becomes consequential.
That makes transparency, monitoring and meaningful human review more important as adoption expands. For job seekers, meanwhile, the incident offers a less comfortable reality: even as AI makes applying for jobs easier, getting an application in front of a person may remain one of the most important steps in the process.
The Logical Indian’s Perspective
The growing use of AI in recruitment can make hiring faster, but efficiency should not come at the cost of fairness or meaningful human evaluation. The reported concerns from a Google DeepMind team highlight why applicants need transparency about how automated screening works.
At the same time, Google’s response shows that such concerns should be examined carefully rather than treated as proof of systemic failure. As AI becomes more involved in hiring, human oversight, accountability and clear processes will remain essential.
Also Read: Maharashtra FDA Suspends Blinkit’s Malad Store License over Cockroaches, Expired Food
