New grads in the US are facing the worst job market since the pandemic as artificial intelligence reshapes hiring
The job market for recent graduates in the United States is going through one of its toughest stretches in decades — and the stories from people living through it are genuinely heartbreaking. The underemployment rate among college graduates hit 42.5% in 2025, the highest level since 2020, according to data from the New York Federal Reserve. That means nearly half of American college grads who finished their degrees are either working in jobs below their qualification level or haven’t been able to break into the market at all.
The number alone is alarming, but what’s behind it is even more revealing about how the world of work has changed — and keeps changing at breakneck speed. Several young graduates shared their struggles with The Guardian, describing what it’s like to navigate a market shaped by shrinking opportunities, the rise of artificial intelligence in hiring processes, and expectations that are increasingly disconnected from the reality of someone just starting out. 🎓
And the landscape has a very specific combination of factors that makes everything even harder:
- Entry-level positions demanding years of experience that no recent grad has
- Hiring processes dominated by artificial intelligence systems that filter resumes before any human ever sees them
- A market that’s hiring less for actual growth and more to replace people who leave
- Companies unwilling to train professionals with related — but not identical — backgrounds
Real stories from people who keep trying — and keep getting ignored
Gillian Frost is 22 years old and studies at Smith College in Massachusetts, majoring in quantitative economics with an additional focus on government. She has been job hunting since last September and described the process as exhausting and frequently demoralizing. According to Frost, every weekend she spends more than two hours just submitting applications. By the time she spoke with the Guardian, she had already applied to more than 90 positions. Of those, about 25% of companies simply vanished without any response, and roughly 55% rejected her automatically.
Even though she managed to land about ten interviews, Frost said the lack of communication from employers was particularly frustrating. Many companies don’t even bother letting candidates know they weren’t selected. She described the feeling as total helplessness, saying that nobody seems to know how to properly prepare for this unique convergence of events. How does someone prepare for a tight job market that coincides with the emergence of AI and the United States’ direct involvement in conflicts? Most previous generations dealt with maybe one of these factors — the current generation is the first to face all three at once.
Then there’s Jeff Kubat, 31, who lives in St. Cloud, Minnesota, facing a different but equally tough challenge. After spending eight years optimizing accounts payable processes at a construction company, he went back to school to get a master’s degree in accounting. Since then, finding a position that matches his qualifications has been a constant battle. According to Kubat, even companies in small Minnesota towns are being incredibly literal about the exact profile they want, showing zero willingness to train people with related experience.
As the search dragged on, Kubat said he was starting to lower his salary expectations. He recognizes that the next job doesn’t have to be the job of a lifetime, but he still needs to pay his bills. And the feeling, he said, is like being stuck in a region that doesn’t match his field. The only openings that seem to pop up are ones created because someone left, not because of genuine company growth — a direct reflection of the numbers showing that hiring has dropped to pandemic-era levels.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Impossible requirements for first jobs
The difficulty isn’t just finding openings — it’s meeting job requirements that keep getting more demanding. A 25-year-old graduate from New York University, with a degree in media, culture, and communications, shared that many positions labeled as entry-level seem completely out of reach. According to her, jobs with reasonable salaries listed as entry-level frequently ask for candidates with three to five years of experience — a timeline that simply can’t be met by someone who just walked off campus.
That same graduate said most job descriptions make her feel so underqualified or insufficiently experienced that she doesn’t even bother applying, since she doesn’t have years of work history to show. And this is a point worth paying attention to: when entry-level positions demand mid-career experience, the entire system breaks down.
This phenomenon has a name among HR experts: credential inflation. It happens when companies artificially raise the requirements for a position without actually needing everything they’re asking for. Some reasons include the surplus of available candidates, the desire to filter application volume without much effort, and in some cases, a real disconnect between what HR requests and what the team actually needs day-to-day. The direct result is that graduates get locked out of the process before they even have a chance to show what they can do — and the real opportunities end up going to candidates who already have prior experience, perpetuating a cycle that’s incredibly hard to break. 😤
When AI decides who gets in and who gets left out
One of the quietest — yet most impactful — changes in the hiring process over recent years has been the widespread adoption of automated resume screening systems. These systems, known as ATS (Applicant Tracking Systems), use layers of artificial intelligence to filter candidates based on keywords, job title history, years of experience, and even resume formatting patterns.
The NYU graduate described the impact of this pretty clearly. For every position, especially those at large organizations that are more likely to use AI in their hiring process, she has to tailor her resume explicitly for that specific role and pack in as many relevant keywords as possible. She described the process as annoying and exhausting, but unfortunately a necessity given the current state of the market and the stage of technological development we’re in.
Her strongest frustration came next: she said she hates having to worry about passing the arbitrary and inscrutable tests of a machine before any human even considers her capabilities and what she could bring to a given role as an individual.
The problem goes beyond personal annoyance. These systems were largely trained on historical data from hires deemed successful — which means they tend to favor profiles that already exist within companies, not fresh profiles like those of recent graduates. In practice, this creates a pretty brutal cycle: the posting asks for two or three years of experience, the system automatically rejects anyone who doesn’t have it, and the candidate never even gets close to an interview. A Harvard Business School study had already flagged this phenomenon before the explosion of advanced language models — and since then, the problem has only gotten deeper.
With the arrival of generative AI tools integrated into these systems, companies now have the ability to process even larger volumes of candidates in fractions of a second. Paradoxically, the process got faster for companies and more frustrating for candidates. The feeling of sending out dozens of resumes and not receiving so much as an automated response has become routine for many young people in the job market.
What makes this dynamic even worse is that many companies don’t even manually review the candidates discarded by their algorithms. A highly qualified profile can be eliminated simply because the resume didn’t use the exact word the system was looking for, or because the formatting confused the ATS parser. This raises a serious question about the quality of opportunities being lost — not just for candidates, but for the companies themselves, which may be letting real talent slip through a poorly calibrated sieve.
Structural barriers and the weight of connections
For Anna Waldron, 22, originally from Portland, Oregon, the structural barriers in hiring practices have made the job search especially challenging. Waldron is about to graduate from Loyola University Chicago in May with a double major in political science and journalism. She shared that she typically applies through platforms like Handshake, LinkedIn, and FlexJobs, but other times she goes directly to companies she knows in Chicago and applies through the careers section on their websites.
What Waldron discovered along the way is that many positions aren’t even posted on those platforms, because companies hire internally or keep the search limited to the company’s existing network of contacts. This makes life much harder for anyone entering the market now who hasn’t yet built a broad web of professional connections.
And the most frustrating part is that Waldron isn’t someone without experience. She completed three internships during college and has skills in both journalism and public policy work, including a stint in the United States Senate. Even so, despite applying to all kinds of positions related to both fields, she still hasn’t been able to land a role. Her case is a clear example of how the problem goes beyond individual qualifications — there’s something structural preventing proven talent from accessing the opportunities that are out there.
What the data reveals about real opportunities
Looking at the data more carefully reveals something important: the problem isn’t exactly a total lack of openings, but rather the nature of the openings available. The American job market has maintained relatively stable employment numbers in absolute terms — but what changed was the composition of those jobs. Hiring for growth positions, where a company expands its team because business is booming, has dropped significantly. What dominates now is replacement hiring, where someone who left gets backfilled.
And in that scenario, the preference tends to go to candidates who can step into the role with as little training as possible — which puts graduates without prior experience at a clear disadvantage. Companies want someone ready to go, and recent grads need a chance to prove themselves. That mismatch is one of the engines driving the current crisis.
Another relevant data point comes from how companies themselves are using artificial intelligence: many of them are leveraging AI not just for hiring, but also to restructure roles internally. This means some entry-level positions that existed five years ago simply don’t exist anymore — they’ve been absorbed by automation or redistributed among more senior employees who now use AI tools to be more productive. The net result is a market with fewer entry points, where opportunities for people just starting out have become both scarcer and more competitive at the same time. 💡
The ripple effect in Brazil and the rest of the world
Although the New York Federal Reserve numbers are American, the phenomenon resonates in a very similar way in the Brazilian context. There, underemployment among young people with college degrees is also a concerning figure. According to Brazil’s IBGE, the unemployment rate among 18-to-24-year-olds with completed higher education remains significantly above the national average — and a good portion of those who are employed hold positions that don’t require a degree.
In Brazil, the adoption of artificial intelligence tools in hiring processes has also grown considerably over the past two years. Large companies, especially in the tech, finance, and retail sectors, already use automated screening systems, gamified assessments with AI-driven behavioral analysis, and even video interviews evaluated by algorithms before any human contact takes place. The process has become more scalable for companies, but it created an additional layer of distance between the candidate and the recruiter — and many Brazilian graduates report the same feeling of invisibility that their American counterparts describe.
On top of that, Brazil faces an extra challenge: inequality in access to quality education means the impact is distributed very unevenly. Opportunities in the national market remain concentrated in major urban centers, and the job requirements at the most sought-after companies keep climbing, creating mounting pressure on those who are just getting started. 🇧🇷
A structural problem, not a passing phase
What’s clear, looking at all of this, is that the crisis facing recent graduates in landing their first jobs isn’t an accident or a temporary blip. It’s the reflection of deep structural transformations in the job market — driven by technology, but also by the organizational and cultural choices companies are making. Automated systems that reject candidates without human review, AI-inflated job descriptions that generate requirements disconnected from reality, hiring that prioritizes replacement over growth, and a widespread reluctance to invest in training new professionals are all pieces of the same puzzle.
The stories of Gillian Frost, Jeff Kubat, Anna Waldron, and the NYU graduate aren’t exceptions — they’re the norm for millions of young people trying to build the beginning of their careers in a landscape that seems designed to keep them on the outside. And while the debate about the role of artificial intelligence in this process remains wide open, those at the starting line of their careers have to deal with this reality right now, understanding the new rules of the game to find viable paths within a system that urgently needs rethinking.
For anyone navigating this landscape, understanding these dynamics is essential. Not because it changes the immediate difficulty, but because it helps you make more strategic choices — about which sectors have real entry-level openings, which technical skills automated systems search for most frequently, and how to position a resume to get past AI filters without losing authenticity. The game has changed, and knowing the new rules is the first step to playing it better.
