Bad news for anyone trying to find a job in today’s AI-driven hiring environment: even Google’s leading AI research division appears to be struggling with automated recruitment systems.
According to an internal HR memo viewed by Bloomberg, Google DeepMind’s AGI Safety and Alignment Team is encouraging applicants to complete a special form in addition to the company’s standard job application. The goal is to ensure that a “real human” actually sees their application.
The document, which carries the warning “PLEASE DO NOT SHARE THIS DOC WIDELY,” was reportedly created by Google DeepMind’s AGI Safety and Alignment Team. The memo acknowledges that Google’s application system could incorrectly screen out qualified candidates or take too long to deliver their applications to the appropriate hiring team.
“We have an applications system with a non-trivial probability your CV will be screened out incorrectly or take too long to reach us,” the memo states. “Filling out this form makes sure that a real human on the team will get to see your application.”
In other words, the AI-powered hiring process appears capable of filtering out applicants who may otherwise deserve consideration, prompting DeepMind employees to establish a workaround that sends applications directly to human team members.
Despite the contents of the memo, a Google DeepMind spokesperson told Bloomberg that the company’s AI-assisted hiring process is functioning properly.
“This team set up a special form to go past the recruiter review, and get their resumes direct to the people on the team,” the spokesperson said. “But there are no shortcuts to getting hired.”
Still, the situation is striking, particularly because the workaround comes from one of Google’s most prominent AI research groups. The fact that members of the AGI Safety and Alignment Team felt the need to create a separate path around the company’s recruitment process highlights the difficulties businesses continue to face when relying on automated systems to evaluate job candidates.
AI has increasingly become embedded in the hiring process, with companies using automated tools to sort applications, screen résumés, and identify potential candidates. But concerns have grown over whether these systems can reliably distinguish qualified applicants from unsuitable ones.
The Google memo adds another uncomfortable dimension to that debate: if even an organization developing some of the world’s most advanced AI systems needs to create a human-review workaround for its own hiring process, questions remain about how reliable automated recruitment systems are for everyone else.
For millions of job seekers who increasingly depend on digital hiring platforms to get their applications in front of employers, the problem could be especially significant. If automated screening systems can mistakenly eliminate qualified candidates, simply submitting a strong résumé may no longer be enough to ensure that a human ever sees it.
