Gen Z Uses AI in Job Interviews Amid Rising Graduate Unemployment
▼ Summary
– The class of 2025 entered a severe job market, with graduate unemployment at 5.7% and underemployment at a high of 42.5%, as the tech sector cut hundreds of thousands of jobs.
– Graduates are increasingly using real-time AI tools and human coaches during job interviews, a service sold by startups like LockedIn AI and Final Round AI.
– Proponents argue it is hypocritical for tech firms to automate hiring and mandate AI use at work while expecting candidates not to use it in interviews.
– Employers are responding by increasing in-person interviews and using assessment methods like whiteboard exercises to counter AI-assisted fraud.
– The core conflict is whether interviews should assess a candidate’s ability with real-world tools or their unaided knowledge, as the system itself promotes pervasive AI use.
Graduates entering the workforce this year are confronting the most difficult entry-level hiring landscape in half a decade. In response, a significant number are turning to artificial intelligence tools for real-time assistance during live video interviews. This practice has sparked a debate over ethics and fairness, while fueling a new sector of startups eager to provide these services. The economic data framing this trend, however, is unambiguous.
Recent figures illustrate a stark reality for new degree-holders. The Federal Reserve Bank of New York reports that unemployment for recent graduates aged 22 to 27 reached 5.7% by the end of last year, notably higher than the 4.2% national average. More telling is the underemployment rate, which tracks graduates in jobs that don’t require their degree. That figure hit 42.5%, a peak not seen since 2020. The technology sector, long a primary destination for top talent, eliminated approximately 245,000 positions in 2025, with an additional 59,000 cuts in the first quarter of this year.
This cohort watched as an older generation was hired, promoted, and then laid off from major firms within a very short timeframe. The perceived lesson was clear: traditional markers of merit and company loyalty offer little security. Armed with this perspective, and after years of academic encouragement to master new technologies, they are deploying AI in the job search as a strategic necessity.
A burgeoning industry has emerged to meet this demand. One startup, LockedIn AI, recently promoted a service named DUO. It provides real-time transcription of interview questions paired with a live human coach who can view the candidate’s screen and offer guidance. While framed as a story of generational adaptation, the announcement functioned as a product launch. LockedIn is not operating alone, its founder is also behind a similar platform called Final Round AI. Questions have arisen about the marketing tactics of these companies, including potentially synthetic online reviews. Technical limitations also exist, as interviewers can sometimes detect when a candidate switches windows to access the aid.
The demand for these tools grows directly from a deteriorating hiring climate. The National Association of Colleges and Employers found that only 45% of employers now rate the job market for new graduates as “fair,” a downgrade from “good” the prior year. Hiring projections are essentially stagnant. When application volumes are high and interview invitation rates can fall below two percent, the pressure to seize any competitive advantage in hiring becomes intense. A Gartner survey notes that 59% of hiring managers suspect candidates of using AI to misrepresent their abilities.
A central argument from proponents hinges on perceived hypocrisy. They point to the very technology firms that often prohibit interview assistance. Google’s CEO revealed that over 30% of new company code is now AI-generated, a significant increase from six months prior. Companies like Amazon and Microsoft actively encourage engineers to use AI coding tools daily. Furthermore, AI-powered applicant tracking systems automatically screen out resumes before human review. The hiring process is heavily automated, except for the candidate’s performance during the interview. For graduates told repeatedly that AI proficiency is critical, being asked to ignore it during a high-stakes conversation feels less like an assessment of skill and more like a test of adherence to an arbitrary rule.
This perspective carries weight but encounters a logical boundary. A distinction exists between using AI to enhance productivity in a technical task and using it to answer questions about personal experience, judgment, and problem-solving. The fundamental purpose of an interview is to evaluate a candidate’s mind and capabilities. Delegating those responses to an algorithm or a remote coach, regardless of the market’s inequities, subverts the core intent of the conversation.
Employers are already adjusting their tactics. Industry data shows the proportion of in-person interview rounds increased from 24% in 2022 to 38% in 2025. Today, 72% of recruiting leaders implement at least one in-person stage specifically to mitigate AI-assisted interview fraud. Many have introduced practical exercises like whiteboard challenges, pair programming, or unstructured discussions that are far less susceptible to real-time technological augmentation.
A more profound question remains: is the traditional interview still the right evaluation method for an AI-driven economy? If the goal is to assess how a candidate works with the tools they will use daily, then banning those tools seems counterproductive. If the goal is to measure innate knowledge and cognitive ability, then AI assistance invalidates the process. Most interviews attempt a blend of both objectives, which is why the current conflict feels so unresolved.
The class of 2025 did not create these conditions. They entered a labour market transformed by post-pandemic corrections, corporate cost-cutting, and a technological revolution that creates and eliminates roles simultaneously. Their use of AI in interviews is not an act of rebellion. It is the rational action of individuals operating within a system that has loudly proclaimed, in nearly every other context, that AI integration is mandatory. The system’s subsequent objection to them heeding that message is a contradiction that demands scrutiny.
(Source: The Next Web)
