Which Companies Allow AI Tools in Job Interviews in 2026

In June 2026, two companies made headlines with opposite decisions about AI in job interviews. Google began piloting an interview format that explicitly invites candidates to use an AI assistant during a coding round. Amazon formalized a policy requiring candidates to acknowledge in writing that they will not use generative AI tools during the process.

That split isn't a temporary tension waiting to resolve. It's the new reality. Companies are making deliberate, durable decisions about AI use in interviews — and those decisions vary enough that going in without knowing where your target company stands is a genuine risk.

Here's what each major policy looks like, why companies landed where they did, and how to find out before you walk in.

Google: AI Is Now Part of the Format

Google's AI-assisted coding interview started as a pilot targeting junior and mid-level software engineering roles on US-based teams. The format is specific: candidates work in a CoderPad environment with a three-panel layout — file explorer, code editor, and a Gemini AI chat panel — over a 60-minute session that moves through bug fixing, core implementation, and optimization.

The AI can respond in the chat panel but cannot directly edit files. That constraint is deliberate. The interview isn't testing who can produce the most output with AI help — it's evaluating whether candidates can frame effective prompts, validate AI-generated responses, and catch errors in code the model produces.

In other words, the skill being assessed is AI fluency: knowing what to hand to the tool, what to verify, and where the model goes wrong. If you have a Google engineering interview later this year, the preparation question isn't just "am I allowed to use AI?" It's "can I demonstrate I know when to trust it and when not to?"

Meta: AI-Enabled Coding, With Plans to Expand

Meta has been piloting what it calls "AI-enabled coding" interviews, in which candidates can access an AI assistant during the coding portion. The company's stated direction is to extend this format across back-end and operations-focused engineering roles throughout 2026.

The logic behind this decision is consistent across companies taking the same position: the interview should simulate the actual job, and the actual job involves AI tools. Hiring someone who performs well without AI but struggles to use it isn't serving the team they're being hired into.

Canva, Shopify, and Rippling: Making It an Expectation

Canva has gone further than simply allowing AI — the company tells engineering candidates directly that it expects them to use tools like Copilot, Cursor, and Claude during technical interviews. This isn't presented as optional. It's framed as alignment with how the company's own engineers work. If you show up to a Canva interview planning to solve everything from memory, you may be misreading the format.

Shopify and Rippling have similarly moved to explicitly welcoming candidates to bring their preferred AI copilots into live coding sessions. The pattern across these companies is consistent: AI fluency is now part of what they're hiring for, and the interview format reflects that.

Amazon: Explicit Ban, Disqualification Risk

Amazon's position is formal and specific. Candidates receive written language requiring acknowledgment of the rule: "To ensure a fair and transparent recruitment process, please do not use GenAI tools during your interview unless explicitly permitted." Using AI tools is categorized as an "unfair advantage" that compromises the authenticity of the evaluation. Violation is grounds for disqualification.

The reasoning is different from companies that allow AI, but it's coherent: Amazon is assessing individual capability that has developed over time. An AI tool that compensates for gaps in that capability doesn't produce useful signal for predicting how someone will perform on the job — it adds noise. They want to know what you can do, not what you plus an AI assistant can do.

Goldman Sachs: Same Position as Amazon

Goldman Sachs has taken the same stance, treating AI use in interviews as an integrity issue rather than a practical one. For roles at Goldman, the expectation is that performance in the interview reflects individual skill — the firm isn't trying to evaluate your AI fluency, and compensating for gaps with AI assistance misrepresents what they'd be hiring.

The Gray Zone: Most Companies Haven't Said

Google, Meta, Canva, Amazon, and Goldman have all made explicit decisions. Most companies have not. A 2026 Greenhouse survey found that only 18 percent of job seekers say employers have clear AI policies for their interviews.

That 82 percent gray zone is where most candidates actually operate. Absent an explicit statement, the default assumption tends to be: probably don't. That's defensible — but it may be more cautious than necessary for companies that simply haven't updated their guidance rather than ones that have actively decided to prohibit it.

The format often tells you something: a timed HackerRank screener in a proctored environment signals a no-tools expectation. A take-home project with a long window and open access signals the opposite. When the format still doesn't clarify, ask.

How to Find Out Before Your Interview

Check the job description and careers page. Companies with explicit policies are increasingly noting them here. Language about AI fluency as a job requirement is a strong signal the interview format accommodates it. Language about "independent problem-solving" or "no external resources" signals the opposite.

Ask during the recruiter screen. "Is the technical round a standard format, or is AI tooling expected?" is a clean, professional question — not a suspicious one. Recruiters have a specific answer for this now, and it removes ambiguity that would otherwise follow you into the session. If the answer is "we don't have a formal policy," that tells you something too.

Read recent Glassdoor and Blind interview experiences. Candidates who've been through the process in the last six months tend to note whether AI was allowed, required, or explicitly prohibited. Anything older than late 2025 is probably stale for this specific question — the landscape moved fast in the second half of 2025.

When the format itself signals openness, you can ask about tools specifically. "The role description mentions AI fluency — does the technical round incorporate that directly?" That's a reasonable question to ask a recruiter or hiring manager, and the answer tells you exactly what to prepare.

How to Prepare Differently Depending on the Policy

When AI is allowed: The preparation shifts beyond standard interview prep. You're not just rehearsing recall under pressure — you're practicing the judgment calls that happen when you have AI available. Which parts of a problem should you hand to the tool? How do you evaluate whether the output is correct? What does it look like to debug AI-generated code that's subtly wrong?

These skills are learnable, but they take practice with the actual tools. Running mock sessions with an AI assistant before the interview — not just using it to answer prep questions — is how you build the muscle.

The candidates who stand out in AI-assisted formats aren't the ones who use the tool the most. They're the ones who can tell when it's wrong and explain why. That's what Google is actually evaluating.

When AI is banned: The standard interview prep applies — but your preparation process can still involve AI. Using an AI assistant to generate likely questions, drill your examples, and identify gaps in your answers is preparation, not assistance. No company policy restricts what you do the night before.

Meeting Copilot's interview assistant handles both scenarios. Before any interview, it builds a briefing pack from your calendar event — the company, the role, the attendees, and the context you load. You walk in prepared regardless of what format you're facing. In companies where real-time AI assistance is permitted, the overlay surfaces suggestions built from your own preparation material, not generic content — the output reflects the context you loaded, so it sounds like you rather than a language model. When the company prohibits live AI, the prep it generates still raises the quality of what you bring to the session on your own.

The Bigger Split

The companies allowing AI in interviews aren't making this decision frivolously. They're betting that AI fluency is now a job-relevant skill, and that assessing it directly produces better hiring signal than assessing recall in a format that doesn't mirror the actual work.

The companies banning it aren't resistant to change. They're making a specific claim: individual capability, measured without AI compensation, predicts performance better than a joint human-AI output. Both bets will be tested as the hiring cohorts from each approach move into their roles.

For now, what you can control is knowing which format you're walking into. Check before you show up, prepare for the specific environment you'll face, and don't prepare for a generic version of an interview that may not exist at your target company anymore.

---

Sources: Northeastern Careers: Google's AI-Assisted Coding Interview 2026 Guide · HR Executive: Google, Meta, Amazon making AI use a job requirement · LockedIn AI: Companies Allowing AI in Interviews · IT Pro: Amazon bans AI tools in job interviews · Greenhouse: 63% of Job Seekers Have Faced an AI Interview · InterviewMan: Which Companies Allow or Ban AI in Interviews 2026