Robert Half Says AI Fluency in Interviews Is Now Baseline, But It’s Also Slowing Hiring

AI fluency in interviews is getting harder to avoid, but easier to misunderstand. Over the last few months, the loudest debate has focused on whether candidates should use AI at all, or whether companies should ban it from resumes, take-homes, and technical screens.

Talent management consulting company Robert Half’s April survey of more than 1,300 U.S. workers points to a more practical answer. In that research, 36% said early-career candidates should be ready to demonstrate knowledge of AI tools, while 37% warned against using AI to overstate skills or experience. The message is clear: AI fluency is now expected, but it is no longer a differentiator.

That distinction matters for anyone interviewing for data, product, analytics, or engineering roles. AI is not only raising expectations for interviews, but also transforms how they move through the overall hiring process, starting with resume screens.

Since early-career candidates are expected to demonstrate AI fluency, interviewers are adjusting what “qualified” actually looks like in practice. Robert Half made the shift explicit in the same survey. It emphasized that employers are not looking for early-career candidates with deep technical expertise. Rather, they are looking for familiarity with AI tools, along with the ability to review AI-generated content, recognize its limitations, and take responsibility for the final product.

That framing lines up with what many hiring teams appear to be optimizing for in 2026. While AI can accelerate drafting, coding, research, and analysis, it cannot remove the need for judgment altogether. A candidate who can prompt a model is useful. But a candidate who can catch a bad assumption, spot a fabricated detail, or explain why an answer is directionally wrong is much more valuable.

As AI adoption increases, the growing ambiguity at the top of the funnel is forcing companies to rethink how they evaluate candidates. A previous survey from Robert Half helps explain why, and how. In a March report of more than 2,000 hiring managers, 67% said AI-generated applications are slowing hiring, as AI-enhanced resumes make skills harder to verify and thus increase hiring teams’ workloads.

Robert Half reported how exactly the use of AI tools among job seekers is extending hiring timelines, considering:

• 42% of teams spend more time reviewing applications,
• 38% are increasing the number of interviews per candidate, and;
• 32% are rewriting job descriptions to discourage generic AI-generated responses

For candidates, this helps explain why so many interview processes feel more conversational, more open-ended, and sometimes more repetitive. The extra steps are not random; they are attempts to get a cleaner read on what a person actually knows.

Interview Query’s own signals, such as coaching sessions and interview experiences from users, align with the same pattern. Interviews for data science, analytics, and product roles still skew heavily toward core skills, including SQL, case, behavioral, and system design screens. Candidates preparing for such roles are thus allocating more prep time on business trade-offs, metric definitions, and recommendation structure.

In other words, AI has not replaced the core interview bar. It has made the bar easier to blur earlier in the funnel, which makes live evaluation more important later on. That is part of why open-ended data science interview questions and realistic scenarios keep gaining weight. They force candidates to show how they reason when there is no perfect template answer.

Candidates who want to train for that kind of pressure usually need practice that feels more like the real loop. That is where mock interviews and live feedback tend to beat passive prep, especially for case and product-style rounds.

Beyond interviews, this shift is starting to change how teams expect candidates to operate on the job. This can be observed in CoderPad’s 2026 State of Tech Hiring report. Based on a survey of more than 650 developers, recruiters, and hiring leaders, the company found that 82% of developers say generative AI is useful in their work. But it also said strong hiring teams are shifting toward assessments that reflect actual job tasks, like debugging AI-generated code, explaining trade-offs, and improving AI output collaboratively.

Those are not AI skills in the narrow sense, as they instead signify working style across organizations. They show whether a candidate treats AI like an answer machine or like a draft partner that still needs scrutiny. That aligns with the broader shift Karat’s 2026 engineering interview trends has previously covered, where more technical loops are moving closer to real work, leading to job-simulation formats.

That distinction is especially important for data candidates. A wrong SQL query, a misleading experiment readout, or a polished but shallow product recommendation can all sound convincing at first. Interviewers increasingly seem to care less about whether a candidate used AI to prepare and more about whether they can catch bad reasoning and address it before it ships.

The practical takeaway is simple. AI fluency should now be treated as baseline professionalism, similar to being comfortable with spreadsheets, docs, or search. However, despite helping someone move faster, it does not replace ownership.

Strong candidates should prepare for interviews in three layers:

• Use AI tools, but be ready to explain exactly what they helped with.
• Review every generated answer as if it will be challenged live.
• Practice turning rough output into a clear, defensible recommendation.

Instead of pretending they never use AI, candidates can stand out by showing that they can use it to be more efficient without completely outsourcing their judgment and critical thinking.

Robert Half’s April survey captured an important shift in one sentence: AI proficiency is becoming a baseline expectation, but judgment and accountability are what define strong early-career performance. Its March hiring data helps explain why. As AI-generated applications flood the top of the funnel, employers are looking for stronger proof deeper in the process.

For interview candidates, that raises the importance of the oldest signals in the room: clear reasoning, honest communication, and the ability to defend a final answer. AI may change how candidates prepare, but it is not replacing the need to think. If anything, it is making that skill easier for employers to test.

• Karat’s 2026 Survey Shows Why Technical Interviews Now Simulate Real Work
• How to Answer Open-Ended Data Science Interview Questions
• The 4 Types of Data Science Interviews (And How to Prep for Each)