The 2026 Hiring Double Standard: Companies Screen You With AI and Won't Let You Use It Back

Goldman Sachs told its campus recruiting applicants in EMEA that they may not use ChatGPT, Google, or any external AI assistance during the interview process. The rule applies under penalty of immediate disqualification.

The same company uses HireVue — an AI-powered video screening platform that scores candidate responses, analyzes speech patterns, and ranks applicants algorithmically — to triage who gets past the initial screening stage.

That is the arrangement. An AI evaluates you before any human sees your name. Then, once a human does evaluate you, you are forbidden from using any AI of your own. Both sides of the encounter are asymmetric: you don't know what HireVue's model is measuring, what flags your response as a mismatch, or how the ranking is calculated. Goldman does.

This Is Not One Company's Policy

Goldman is the most-cited example, but it isn't alone.

Amazon formalized a policy requiring candidates to acknowledge in writing that they won't use generative AI tools during interviews, categorizing AI use as an "unfair advantage" that compromises the authenticity of the evaluation. Industry reporting places Amazon among HireVue's enterprise clients alongside JPMorgan, Citi, Bain, BCG, IBM, and Capital One — companies that use the AI screening platform at scale to manage candidate volume.

Anthropic — the company that builds Claude — told job seekers last year that they cannot use AI to fill out job applications, because the company wants to see "genuine, unmediated interest." It is one of the more specific ironies of the current moment: the team building some of the most capable AI in the world doesn't want candidates demonstrating how they'd use it.

The pattern is consistent. Companies that have made AI central to their own operations — Goldman's internal AI platform, Amazon's tooling deployed across tens of thousands of employees, Anthropic's core product — are simultaneously telling candidates that AI assistance represents an unfair advantage they cannot use. On one side of the hiring table, AI is a productivity multiplier. On the other, it's cheating.

What "Fair" Means When One Side Has Better Information

The fairness argument for banning candidate AI use has an internally coherent logic. Companies want to assess what an individual can do independently — not what a human-AI team can produce in a controlled session. For roles where individual problem-solving under pressure matters, testing it directly makes sense.

The logic becomes harder to defend when you place it next to how most candidates are actually evaluated before they reach a human interviewer.

The Greenhouse 2026 Candidate AI Interview Report found that 63% of active job seekers have been interviewed by an AI, up 13 percentage points from just six months earlier. Seventy percent of those candidates were never told upfront that an AI was evaluating them. Twenty-one percent only found out after the process had already started.

So the sequence, for a typical candidate at a screening-heavy company, looks like this: submit an application that gets ranked algorithmically against the job description, pass a video screening evaluated by AI you may not have known was AI, and then arrive at a human interview where you are explicitly prohibited from any AI assistance. At every stage except the last one, you have incomplete information about the criteria you're being judged by — no rubric, no threshold, no feedback when you don't advance.

This asymmetry is what has moved the conversation from an industry debate to mainstream business press. Harvard Business Review ran a piece in June 2026 titled "AI Has Broken Hiring." A Futurism headline from earlier this year captured the temperature: "Companies Bragging About Their AI Furious as Job Applicants Use AI During Interviews."

The frustration isn't primarily about whether AI use during interviews is cheating. It's about who gets to define what information each side of the table can access.

The Candidate-Side Numbers

Fifty-nine percent of hiring managers in a 2026 survey reported suspecting candidates of using AI tools to misrepresent their abilities during live assessments. That suspicion has driven the formalization of bans — a preventative response to a perceived integrity problem.

What the same research tends not to surface: only 8% of candidates in the Greenhouse study believed the overall AI-driven hiring process was fair to them. Forty-six percent said they want the option to request a human interviewer. Forty-four percent want upfront disclosure when AI is evaluating them — a transparency standard that would apply to any other third-party assessor.

Companies and candidates are not simply disagreeing about whether candidate AI use is problematic. They have opposite views about where the core fairness problem sits. Companies see the integrity risk in what candidates might use. Candidates see the transparency risk in what's being used on them.

New Laws Are Starting to Close the Gap

The asymmetry has attracted legislative attention. The laws coming out of it are aimed at employers, not candidates.

Illinois amended its Human Rights Act, effective January 1, 2026, to require employers to disclose the use of AI for any employment-related decision — hiring, promotion, discipline. The amendments also prohibit AI use that results in discrimination on protected characteristics, and require employers to preserve records of that use for four years.

Colorado went further. SB 205 took effect June 30, 2026. Under the Colorado law, companies using AI to make or "materially influence" employment decisions must notify candidates before AI is used in their evaluation, conduct and publish annual impact assessments, provide candidates with a right to human review of adverse AI-driven decisions, and publish a public statement describing the AI systems in use and how they manage discrimination risk.

New York City already requires annual bias audits for AI hiring tools, with public disclosure requirements. California, Maryland, and Texas have related legislation at various stages of the process.

The practical effect for candidates in covered states: you now have legal rights you didn't have two years ago. In Colorado and Illinois, employers who use HireVue, resume screening algorithms, or any other AI system to make hiring decisions must tell you. In Colorado, if you don't advance because of an AI-driven decision, you have the right to request a human review.

These laws don't cover every employer or every state, and enforcement will take time to establish. But the direction is now clear. The era of undisclosed AI screening is narrowing, and companies that haven't updated their practices are on notice.

What to Do With This Right Now

The regulatory landscape is shifting. The practical reality of job searching in 2026 is not.

Know the specific rules before you walk in. Companies with formal AI bans — Amazon, Goldman Sachs — are explicit about them. The majority of companies haven't stated a policy one way or the other. Going in without checking is avoidable risk. A direct question to your recruiter — "Does the interview format have any guidelines around AI tool use?" — is professional and removes ambiguity you'd otherwise carry into the session.

Nothing restricts how you prepare. Every company's no-AI-during-interview policy covers the session itself. None of them prohibit using AI the evening before to generate likely questions, practice your answers aloud, identify gaps in your knowledge of the company, or build a detailed briefing on the role and the interviewers you'll meet. Preparation isn't assistance. The distinction is real and it matters.

The human interview is where the outcome is determined. AI screening is a gate. Clearing it gets you into a conversation with a person — and that conversation is what produces the offer. Companies that invest in AI screening at scale are often less prepared for candidates who show up genuinely ready for a real discussion. The preparation gap between candidates who took the human round seriously and those who treated it as an afterthought is wide, and in competitive markets it's decisive.

Meeting Copilot's interview assistant is designed for that window. Before the call, you load the job description, your resume, and any research on the company and role. During the interview — whether AI tools are allowed or not — the preparation you built surfaces when you need it. Your examples, your company context, the things you spent time developing come up in the moment rather than sitting in a browser tab you can't check. That's not an AI doing the interview. It's your own preparation, organized so it's actually usable.

Where This Is Going

The debate has two legitimate positions, and neither is going away.

Companies banning candidate AI are making a claim about what they're measuring: individual capability developed over time, not augmented performance under observation. For roles where that distinction matters, the argument holds.

Companies like Google, Meta, Canva, and Shopify — who explicitly encourage or require AI use during interviews — are making the opposite bet: AI fluency is itself a job-relevant skill, and testing it directly produces better signal than testing recall in a format that doesn't reflect the actual work.

What is harder to sustain is the current middle ground: using AI extensively to screen candidates while prohibiting candidates from any reciprocal use, without disclosing either side of the process. Colorado and Illinois have made a legal judgment that this asymmetry is unacceptable. The pattern of legislation suggests that judgment will extend to more jurisdictions.

For candidates in 2026: understand the specific rules at each company you're interviewing with, prepare for the human conversation like the high-stakes, rarely-won opportunity it actually is, and pay attention to the regulatory landscape. Your rights in this process are changing — and they're changing in your direction.

---

Sources: Fortune: Goldman Sachs Wants Students to Stop Using ChatGPT in Interviews · eFinancialCareers: Goldman HireVue · Greenhouse: 63% of Job Seekers Have Faced an AI Interview · Hinshaw Law: Illinois AI Employment Regulations 2026 · AI Laws by State: Colorado SB 205 · The Employer Report: Colorado and Illinois AI Transparency · HBR: AI Has Broken Hiring