Amazon AI interview policy and 4 other employers

What is Amazon's AI interview policy?

Amazon tells candidates not to use generative AI in interviews unless it is allowed, according to internal guidelines Business Insider reported on February 27, 2025. "To ensure a fair and transparent recruitment process, please do not use GenAI tools during your interview unless explicitly permitted," the guidelines say, as quoted by Fortune.

"Failure to adhere to these guidelines may result in disqualification from the recruitment process," they add. An Amazon spokesperson, Margaret Callahan, told CNBC in March 2025 that Amazon asks candidates to acknowledge they won't use unauthorized tools during the interview or assessment process. That is the closest thing to an on-record statement we found.

We found no candidate AI rule on Amazon's own How We Hire pages when we checked them on October 22, 2026. What Amazon has published is its own use of AI. A January 22, 2026 article on its news site says its new coding assessments let candidates "work with an interactive AI assistant on a real coding platform" that mirrors how its engineers work. So at Amazon, AI is allowed where the assessment provides it and banned elsewhere.

What do the five employers' policies say?

Two of the five publish a candidate AI policy on their own careers pages: Accenture and McKinsey. Amazon's rule is known from news reports and a spokesperson's statement. Google and Meta have published no candidate rule we could find, and news reports say both are piloting interview rounds where AI is provided.

Google's and Meta's careers pages load much of their content with scripts, so a page we couldn't fully read may still hold a rule. If you're interviewing there, ask your recruiter directly. The date column shows when each row was last checked.

What does Accenture's policy allow and ban?

Accenture's page, titled "Redefining hiring with the power of responsible AI," allows AI for research, resume wording and interview practice, and bans it for live answers. "Use AI to rehearse common questions, get feedback, and boost your confidence," it says under practice. Then it adds that when it's time to interview, Accenture wants to hear from you.

The page lists four uses as not permitted. They are false or misleading experience on a resume, AI to complete assessments "unless explicitly stated otherwise," AI products that generate interview answers, and voice cloning or deepfake tools. It says these "can result in disqualification" and that Accenture uses "proctoring tools and AI-detection methods" to keep the process fair.

Accenture also says how it uses AI on its side. The page says AI may help match candidates to roles, auto-fill application forms and highlight how resumes match postings. "You can choose whether you consent to AI being used as part of your evaluation," it adds. What those detection methods can and can't catch is covered in can interviewers detect AI copilots.

What does McKinsey's policy say, including about recording?

McKinsey publishes a six-row table on its Interviewing at McKinsey page. Polishing a resume, practicing interview questions and asking AI to explain concepts during prep are marked encouraged. Inventing achievements, generating real-time interview answers and using AI in assessments "unless specifically permitted" are marked not allowed.

"Using AI to misrepresent yourself or to generate real-time responses during an interview goes against our values," the page says. Its tips for virtual interviews also tell candidates to "Disable previously installed AI software (such as note-taking tools or virtual assistants)" because these may record or assist with the interview.

McKinsey's separate Assessment Integrity Expectations page goes further. Candidates are expected to refrain from "recording any portion of their assessment/interview or taking screenshots" and from using "generative AI, a calculator, or prepared notes" during it. Breaking these "may result in disqualification from our process, rescinding of an offer, restrictions on future applications, or legal action," the page says.

What have Google and Meta said about AI in interviews?

Neither has published a candidate rule we could find, and both have been reported moving AI into some rounds. A Google spokesperson confirmed a pilot for some software engineering interviews to Business Insider in May 2026. Meta's test of AI-assisted coding interviews was first reported by 404 Media in July 2025.

At a February 2025 town hall, Sundar Pichai said it was "worth thinking about some fraction of the interviews being in person," CNBC reported. Business Insider then reported in May 2026 that Google will let candidates use an "approved" AI assistant, which a spokesperson said would be Gemini, in a code comprehension round. Interviewers will evaluate "AI fluency, including prompt engineering, output validation and debugging skills," according to the internal document Business Insider reviewed.

At Meta, the new format "is more representative of the developer environment that our future employees will work in," according to the internal post quoted by 404 Media. Neither company has said what happens to a candidate who uses an unapproved tool. For how other employers changed their process, see how companies counter AI copilots.

Why do the same companies ban AI in one round and provide it in another?

Because the rule is about who chose the tool and whether the interviewer can see it. A provided assistant is part of the test, and the interviewer watches how you use it. A hidden one answers for you, so the interviewer can no longer tell what you know. Every published or reported position from these five companies fits that split.

Amazon's own example shows it. Its January 2026 article describes AI built into coding assessments, while its reported guidelines ban GenAI tools in interviews unless they are explicitly permitted. McKinsey marks AI in assessments "Not Allowed" unless "specifically permitted," and a January 2026 Fortune report says some McKinsey candidates were asked to work with Lilli.

So "is AI allowed" has no single answer at any of these companies. It depends on the round, and the employer decides. The page on where using AI in an interview becomes cheating sets out the before, during and after line that holds across all five.

What should you do before an interview at one of these companies?

Ask your recruiter two things in writing: whether any AI tool is allowed or provided in any round, and whether recording or AI note-taking is allowed. Get the answer before the interview, follow it exactly, and keep the email. Rules change by round, by team and by date, and the recruiter's answer is the one that applies to you.

The second question matters more than it looks. McKinsey's own page warns that AI note-taking tools "may unintentionally record or assist with the interview," and its integrity page asks candidates not to record. Amazon's own interview transcription pilot says it requires "explicit consent from candidates before any interview recording begins," so consent runs both ways. The script for how to ask to record your interview works for that conversation.

Prept's line is analysis after the call, never during it, and before every recording it asks you to confirm that everyone on the call knows. Where an employer bans recording, as McKinsey does, use the free 10-minute written debrief template as soon as the call ends instead. If you're unsure what counts as allowed help, AI interview help that isn't cheating covers specific tools.