AI in hiring: what's actually changing, and how to prepare

Employers now use AI to screen resumes, and some to run interviews. Here's where it shows up, what it tests, the risks, the new rules, and what candidates should do differently.

Both sides of the hiring table now use AI. In a May 2026 Novoresume survey of 2,000 US adults, 42.6% said they used AI tools the last time they updated their resume. On the employer side, SHRM's 2026 research found 27% of organizations use AI in talent acquisition, more than any other HR function. Among organizations with 5,000 or more employees, 60% use AI somewhere in HR.

That combination creates a lot of anxiety. Candidates picture a robot rejecting them, and employers picture a flood of AI-written applications. The reality is more specific, and knowing the specifics helps you prepare.

Where employers use AI today

According to SHRM, the most common uses in recruiting are writing and optimizing job postings, screening and matching resumes, and scheduling interviews. In other words, most of the AI you meet as a candidate is sorting and logistics, not judgment.

That's starting to change with AI interviewers. HackerRank, a technical hiring platform, launched an AI interviewer called Chakra in October 2026 after more than 500,000 interviews in beta at companies including Snowflake and Capgemini. Chakra scores candidates, but people still make the hiring decision.

What AI interviewers actually test

The interesting part is what these tools measure. Chakra gives candidates realistic problems to solve with AI assistance, then asks about their choices, their reasoning and whether they checked the AI's work. HackerRank calls this AI fluency.

On a recent episode of the Possible podcast, Aria Finger described it as a sign of where hiring is going: employers increasingly need to know how a candidate directs and evaluates AI, because, in her words, everyone is going to be a manager of AI. That applies well beyond engineering. Marketers, product managers and analysts are all being asked some version of the same question: can you get good work out of these tools, and can you tell when they're wrong?

The optimistic case

Reid Hoffman, LinkedIn's co-founder, made the case for AI in hiring on the same episode. "I actually think AI inclusion in the hiring process will broadly be much better," he said. His argument is about cost. Human interviews are expensive, so employers narrow the funnel with crude filters such as requiring a bachelor's degree. If a thorough AI-led skills assessment costs very little, companies can afford to evaluate more people on what they can actually do.

He also expects a division of labor rather than full automation. AI assesses competence, capabilities and learning curve. People assess team and culture fit and cross-check the AI. For ongoing team roles, he doesn't think hiring should ever be handed entirely to AI, because a team needs to buy into its new colleague. The one exception he offered was short contract work where all you need is the finished task.

The risks

The obvious risk is bias at scale. A biased interviewer affects the candidates they meet. A biased screening model can affect everyone who applies through it. That concern is now being tested in court. In Mobley v. Workday, applicants allege that Workday's screening tools discriminated against them based on age, race, sex and disability. The court allowed the claims to proceed on the theory that Workday acted as an agent of the employers using it, and in May 2025 it granted preliminary certification of a collective action on the age claim. Workday denies that it makes hiring decisions, and the case is still pending.

There's also a cost on the candidate side. In the same Novoresume survey, 74.8% of people who let AI write most or all of their resume admitted to something inaccurate on it, compared with 24.6% of people who used no AI. The survey was run by a company that sells resume tools, so treat the exact figures with care, but the pattern makes sense: AI happily fills gaps with claims you can't back up, and interviews are designed to find them.

The rules are catching up

Regulators have started to set guardrails. Since July 2023, New York City has required employers using automated employment decision tools to have them independently audited for bias each year, post a summary of the results, and tell candidates at least 10 days in advance, including that they can ask for an alternative process or an accommodation.

In California, regulations that took effect on October 1, 2025 make clear that using an automated decision system that discriminates against applicants or employees can violate state civil rights law, even when the tool looks neutral. Employers must keep those records for four years, vendors acting on an employer's behalf can be treated as its agents, and evidence of anti-bias testing before and after adopting a tool can help an employer's defense.

How to prepare as a candidate

Practice working with AI out loud. If an interview lets you use AI, the evaluator is watching how you frame the problem, what you ask for, and whether you catch mistakes. Narrate your reasoning, check the output against what you know, and say when you'd override it.

Use AI to draft, never to invent. Let it help you tighten bullets and tailor a cover letter, then check every claim against what you actually did. If you can't defend a line in an interview, delete it.

Treat an AI interview like a real one. Set aside the full time, find a quiet space and prepare examples with numbers, exactly as you would for a person. If you need an accommodation, ask for it. In New York City, you're entitled to be told how to request one.

Debrief after every interview. On the podcast, Aria described someone who had done 20 first-round interviews without getting feedback on any of them. Write down the questions and your answers right after each one, then ask an AI tool where you were strong and where you rambled. If you'd like to record a call instead, get permission first. California and several other states require everyone on a call to consent.

Keep a human path. AI makes it cheaper to apply and cheaper to screen, which means more applications competing for the same roles. A referral or a note to the hiring manager still stands out, and as our post on job search myths covers, referred candidates are hired at far higher rates than cold applicants.

The bottom line

AI is changing how candidates are found and assessed, but the fundamentals haven't moved: apply to roles you genuinely fit, tell the truth well, show your judgment, and get a person on your side. Plumrole is built around that: it ranks live roles by fit, drafts tailored materials only from your real experience, and leaves every decision to you.