Google DeepMind warns job applicants about AI screening risks

Google sells AI tools to companies to make hiring faster. But one of its own AI teams is warning candidates about the risks of automated screening.

According to Bloomberg, Google DeepMind's AGI Safety and Alignment Team has asked candidates for some roles to complete a special form with their applications.

The move aims to prevent CVs from being wrongly screened out or delayed by Google's internal hiring systems.

Google team bypasses recruiter screening

The team works on reducing risks linked to advanced AI.

A document seen by Bloomberg asks candidates to submit an additional form. The document was marked "PLEASE DO NOT SHARE THIS DOC WIDELY".

It states, “We have an applications system with a non-trivial probability your CV will be screened out incorrectly or take too long to reach us,”

It adds, “Filling out this form makes sure that a real human on the team will get to see your application.”

The team also warned candidates against relying on AI for their applications.

It said human reviewers were tired of generic large language model (LLM) responses.

The advisory reads: “A real human will read these. These humans get really tired of reading LLM (large language model) answers, because they all sound very samey.”

Google DeepMind disputed the suggestion that its systems wrongly filter candidates.

A spokesperson said the company aims to “recruit and hire the most qualified talent at Google DeepMind”.

The spokesperson said the form simply allows the team to receive CVs directly. It bypasses the standard recruiter review.

“This team set up a special form to go past the recruiter review, and get their resumes direct to the people on the team. But there are no shortcuts to getting hired,” the spokesperson added.

AI is changing recruitment

AI is becoming a bigger part of hiring.

Companies use it to draft job descriptions and analyse CVs. Some systems also rank applicants and forecast hiring needs.

Google promotes its AI tools to businesses for these tasks. Its Workspace products include features designed to help HR teams evaluate resumes and plan recruitment.

But automated hiring has raised concerns.

Some companies use AI to rank candidates. Others filter CVs using keywords. These systems can create a risk of qualified candidates being overlooked.

Candidates are gaming AI systems

Applicants are also using AI to speed up job searches.

Some are using AI to produce CVs and cover letters in large numbers. Others are trying to manipulate automated screening tools.

A 2025 report by The New York Times found that some candidates had hidden instructions inside CVs. These prompts were designed to influence AI screening systems.

The trend creates a new problem for recruiters.

AI can process large numbers of applications. But it can also make genuine responses harder to identify.

AI hiring faces bias concerns

AI recruitment systems are also facing scrutiny over potential bias.

A US federal court allowed claims to proceed against Workday in the class action Mobley v. Workday.

The lawsuit alleges that Workday's AI-powered recruitment software discriminated against applicants based on race, age and disability.

Workday has denied the allegations. It says its systems support responsible hiring.

Research has raised similar concerns.

A July 2025 academic paper, "Fairness Is Not Enough: Auditing Competence and Intersectional Bias in AI-powered Resume Screening", examined AI-powered CV screening.

The research found that some systems could rely on superficial signals. Others could depend heavily on keyword matching.

Human judgement remains important

The Google DeepMind case highlights a key challenge for AI-powered recruitment.

Companies want faster hiring. They also want better candidate screening.

But AI systems can make mistakes. They can also create bias and be manipulated by applicants.

The challenge is finding the right balance between automation and human judgement.

For candidates, the lesson is similar. AI can help prepare an application. But it cannot guarantee that the CV will reach the right person.