AI and Entry-Level Jobs: The Real Data for College Graduates 2026

Last Updated: July 2, 2026

Entry-level job postings in the US are down 35% since early 2023. Unemployment among recent college graduates (ages 22-27) sits at 5.7% - well above the 4.3% rate for all workers. And 43% of new graduates are now underemployed - the highest rate since the peak of the pandemic. These are real numbers from real data sources, and they describe real pain for the graduating class of 2026.

Here is the counterintuitive part: since ChatGPT launched in late 2022, the US economy has added approximately 3 million white-collar jobs. There are 7% more software developers than there were in 2022, 10% more radiologists, and 21% more paralegals. The overall unemployment rate remains near historic lows.

Both sets of facts are true simultaneously. Understanding how requires looking at something more specific than "AI is taking jobs" - a framing that is both accurate and deeply misleading depending on which part of the job market you are looking at.

🎯 Before you read on - we put together a free 2026 AI Tools Cheat Sheet covering the tools business leaders are actually using right now. Get it instantly when you subscribe to AI Business Weekly.

Get the Free Cheat Sheet

The data on entry-level hiring for recent college graduates paints a picture that is worse than the top-line employment statistics suggest - and better than the most alarming AI predictions.

The hard numbers:

Entry-level job postings in the US are down 35% since early 2023, according to labor market tracking data cited by Metaintro's April 2026 labor market analysis. Handshake found that job postings to their platform aimed at college students and recent graduates dropped 16% even as applications per opening jumped 26% - meaning more competition for fewer spots, per US News opinion by Jane Swift and Ryan Craig.

Unemployment among recent college graduates ages 22-27 reached 5.7% in late 2025 and remains elevated - well above the 4.3% rate for all workers. "This disparity in unemployment is not something that throughout history has been true all that often," Cory Stahle, senior economist at Indeed, told Forbes in May 2026.

Nearly 43% of new graduates are now classified as underemployed - the highest rate since the height of the pandemic - per Metaintro. Underemployment means working in jobs that do not require a college degree, or working part-time when full-time work is wanted.

The share of long-term unemployed workers with bachelor's degrees is now approximately 1 in 3, per Washington Monthly's May 2026 analysis of the entry-level market. A decade ago, it was 1 in 5. That shift reflects something structural, not cyclical.

The morale data:

Job market confidence has deteriorated sharply. Only 19% of college graduates say it is a "good time to find a quality job" - versus 35% of those without degrees, per Yale's Jeffrey Sonnenfeld and co-authors writing in Fortune, referenced in Yale Insights. That is a historical reversal. A college degree has historically correlated with more economic optimism, not less. The gap between how graduates feel and how the top-line statistics look is the central tension of the 2026 labor market.

For broader data on how AI is affecting workplace productivity and employment, our AI adoption statistics guide covers the full enterprise picture.

The seeming contradiction - 3 million new white-collar jobs since 2022 alongside a contracting entry-level market - resolves when you look at where those jobs are and who is filling them.

Most of the white-collar job growth has occurred in roles that require demonstrated experience, domain expertise, or technical credentialing - not just a diploma. Software developers are up 7%, but the roles driving that growth skew toward senior engineers, AI specialists, and cloud architects, not junior developers. Radiologists are up 10%, but those positions require medical degrees and residencies. Paralegals are up 21%, but many of those gains reflect experienced workers expanding their capacity with AI tools rather than new graduates entering the field.

The pattern is described precisely by Washington Monthly: AI is not wiping out job categories. It is wiping out the bottom layer of those categories - and the bottom layer is where new graduates historically started.

Think about what a junior analyst actually did in 2019. They pulled data, cleaned spreadsheets, drafted first versions of reports, summarized research, made calls, and did the operational groundwork that senior people needed done before they could apply their judgment. That work - valuable, learnable, and perfect for someone a year out of school - is increasingly the work that AI tools handle now. A senior analyst still exists, still earns more than before, and now uses Claude or ChatGPT to do what the junior analyst used to do. The junior role does not disappear from the company's org chart immediately. It just stops being filled.

Stanford researchers found that workers aged 22-25 in jobs most exposed to AI - software development, customer service, and accounting - experienced a 16% relative drop in employment in under three years, per US News. The same AI exposure that increases the productivity and job security of experienced workers is decreasing the demand for inexperienced ones in identical roles.

The most useful concept for understanding what is happening to entry-level hiring in 2026 was coined by Laura Ullrich, lead economist at Indeed: "experience creep."

Experience creep describes a pattern where employers demand more experience for jobs that once were available to candidates still acquiring it. It predates AI - employers have been gradually inflating experience requirements for years. But AI is accelerating it dramatically.

Here is how it works in practice. When AI tools can handle the basic analytical, writing, and research tasks that entry-level employees traditionally used to do, employers no longer need someone to do that work. What they need is someone who can manage the AI doing that work, evaluate its outputs, catch its errors, and apply judgment about what to do with the results. That skill set requires experience. You cannot develop it in a classroom.

The result, as Washington Monthly puts it precisely: "AI is elevating employer expectations for new employees. Graduates are expected to perform higher value work right out of the gate - a nearly impossible task when so many lack prior relevant experience."

The entry-level job's traditional function was not just to provide labor. It was to be the place where graduates developed the skills, judgment, and professional context that would make them valuable in five years. When that on-ramp disappears, the graduates who cannot access it fall behind in ways that compound over time - not just in their first year, but in their career trajectory.

This is why Yale's Sonnenfeld and co-authors describe the AI employment problem as one that is "hitting before careers can start." The wipeout of experienced knowledge workers that AI pessimists predicted has not arrived. The quiet disappearance of the on-ramp to becoming an experienced knowledge worker - that is the part that is real and measurable right now, per Yale Insights.

💡 Finding this helpful? Get bite-sized AI news and practical business insights like this delivered free every morning at 7 AM EST.

Subscribe Free

Not all graduates are experiencing the same market. The data shows meaningful variation by major - including some results that run counter to conventional wisdom.

The counterintuitive finding on computer science:

The Federal Reserve Bank of New York found that unemployment among recent college graduates in computer science is 7.0% and in computer engineering is 7.8% - among the highest of all majors and comparable to rates for anthropology, fine arts, and performing arts, per Yale Insights.

This surprises people who assume that CS graduates are insulated from AI displacement. The reality is more nuanced: the junior software developer role - the traditional entry point for CS graduates - is one of the most directly affected by AI coding tools. Claude Code, GitHub Copilot, and Cursor have dramatically reduced the demand for developers who can write boilerplate, debug basic errors, and build standard features. Senior developers using those tools now do what junior developers used to do. The junior roles get cut.

The exception is at the frontier of AI itself: IBM tripled entry-level hiring specifically targeting software development, cybersecurity, and AI engineering, per Metaintro. The demand for people who can build and deploy AI systems is growing. The demand for people who can do the tasks AI systems now handle is shrinking.

The majors facing the most exposure:

Stanford's research identified software development, customer service, and accounting as the roles where workers aged 22-25 have experienced the steepest relative employment decline. Per US News, basic data entry, administrative work, and basic coding are "precisely the roles AI tools are absorbing fastest."

The fields with more resilience:

Healthcare, cybersecurity, public sector, and skilled trades have maintained or increased entry-level hiring. These fields share common characteristics: physical presence requirements that AI cannot substitute for, regulatory frameworks that require human accountability, or genuine labor shortages that technology has not yet addressed.

Critical thinking, complex problem-solving, adaptability, and creativity rank as the most sought-after capabilities by employers in 2026 - by a wide margin over technical execution skills, per Yale Insights. This is a real shift from five years ago, when employers valued execution competency first.

Monster's 2026 Graduate AI Readiness Report - based on more than 1,000 recent and upcoming college graduates surveyed in January 2026 - is the most comprehensive direct data on how graduates are experiencing this moment, per Monster's report:

The commencement speaker data from Washington Monthly adds texture that surveys cannot: "College graduates' dread about the job market is so acute that commencement speakers who dare mention artificial intelligence keep getting booed." The jeers are not panic. They are exhaustion - from four years of being told AI would either destroy their careers or supercharge them, without being given the specific tools to navigate either outcome.

The most actionable finding in the 2026 data is not about AI capabilities or labor market trends. It is about a structural failure in how American higher education prepares students for work.

The internship gap is severe:

ZipRecruiter's 2026 Annual Grad Report found that graduates with work experience during college were hired at a rate of 81.6%, compared to just 40.7% for those without - a more than two-to-one difference, per Metaintro. Internships and part-time roles in related fields are no longer differentiators - they are filtering mechanisms. The market is dividing graduates into those who have demonstrated professional competency before graduating and those who have not.

The co-op gap is worse:

Fewer than 0.5% of American college students participate in formal co-op programs - the structured alternating work-study arrangements that are standard in Canada, where nearly all large and mid-sized universities offer four-month academic terms alternating with four-month in-field paid work, per US News. American higher education's resistance to mandatory work experience creates graduates who are academically credentialed but professionally untested - exactly the combination that AI tools most directly displace.

The AI curriculum gap:

Only 36% of graduates say college is preparing them to use AI professionally. 47% of recent graduates say AI has already impacted hiring in their field, but only 23% of those grads received extensive AI training during their studies, per Metaintro. The gap between AI's presence in the job market and AI's presence in the curriculum is the most urgent gap in American higher education right now.

The experience paradox:

As Washington Monthly concludes: "Getting their first job now requires proof that they have already had one." That paradox - the experience requirement for a job that is supposed to provide the experience - has always existed to some degree. AI has made it acute. When AI tools handle the basic work that entry-level jobs used to provide as on-the-job training, the market's tolerance for inexperience at entry narrows sharply.

The uneven distribution of entry-level hiring in 2026 is the most practically relevant data for graduates making decisions right now.

Healthcare:

Healthcare continues to hire entry-level graduates at meaningful volume. The sector's requirements for physical presence, direct patient interaction, and regulatory accountability create roles that AI cannot substitute for at any near-term price. Social work, hospital administration, patient advocacy, public health, and health informatics all represent genuine opportunities. The caveat: administrative roles within healthcare that involve primarily data processing and documentation are facing the same AI pressure as their counterparts in other industries.

Cybersecurity:

IBM's tripling of entry-level cybersecurity hiring is representative of a broader trend. The demand for people who can defend AI-enabled systems, manage security incidents, and audit AI deployments is growing faster than the supply of people who can do it. This is one of the clearest cases where AI creates direct new entry-level demand by creating new attack surfaces and compliance requirements.

AI engineering and deployment:

The distinction between AI as a job displacer and AI as a job creator resolves clearly here: building, deploying, fine-tuning, evaluating, and maintaining AI systems requires human engineers. Entry-level roles in model evaluation, prompt engineering, AI safety, dataset curation, and LLMOps have emerged as genuine career on-ramps that did not exist three years ago. For CS graduates specifically, pivoting toward AI engineering rather than general software development is the clearest path to an expanding rather than contracting entry-level market.

Government and public sector:

Federal, state, and local agencies continue hiring in social services, policy analysis, infrastructure, and program management. These roles are often less visible on major job boards than private-sector postings but represent steady volume. Government hiring has been more insulated from AI-driven elimination than the private sector - in part because accountability requirements make automation adoption slower and more deliberate.

Skilled trades and logistics:

Construction, electrical work, HVAC, and supply chain operations are experiencing genuine labor shortages with no AI solution on the horizon. Management-track roles for college graduates are increasingly available to those willing to start on the operational floor. This path runs counter to most graduates' expectations but offers genuine career development in an undersupplied talent market.

For broader data on how AI is affecting employment across industries, our AI productivity statistics guide covers the full enterprise landscape.

ZipRecruiter's 2026 Graduate Report and reporting from multiple labor economists converge on a consistent profile of the graduates successfully navigating this market.

They have demonstrated work experience before graduating:

The 81.6% vs 40.7% hiring rate difference between graduates with and without work experience is the most actionable data point in the entire 2026 labor market picture. The graduates landing jobs overwhelmingly had internships, co-ops, part-time roles in their field, or significant project-based work they can speak to concretely. The diploma is the prerequisite, not the differentiator.

They have genuine AI tool fluency - not awareness, fluency:

The 69% of graduates who believe AI skills will give them an advantage are right. But the advantage does not come from having used ChatGPT to write essays. It comes from being able to demonstrate in an interview how you use AI tools to produce professional outputs - how you prompt effectively, evaluate quality, catch errors, and integrate AI into a workflow that produces better results than working without it. Employers are increasingly asking candidates to demonstrate AI use in real time during interviews.

They treat networking as infrastructure, not networking:

ZipRecruiter's data shows candidates landing jobs are "leaning hard on internship experience, AI fluency, and networking over applications alone," per Metaintro. In a market where job postings are down 35% but companies are still hiring, a meaningful share of entry-level roles are being filled through referrals, warm introductions, and direct outreach rather than application portals. Applying to 200 postings cold is less effective than having someone internal advocate for your candidacy.

They keep an open mind about what the first job looks like:

As financial services hiring manager Duncan told Forbes: "Most financial advisors start without much experience and they gain it by getting trained at a company." The graduates landing jobs often entered fields through adjacent paths - a psychology major landing in financial services, a journalism major landing in content strategy at a tech company. The first job does not have to be the ideal job. It has to be the job that starts the clock on professional experience.

They translate college experience deliberately:

Career experts consistently advise graduates to identify the transferable activities from college - "Did you get to design something? Did you get to build something? What was the problem you were trying to solve?" - and translate those into the language of professional competency rather than academic achievement, per Forbes.

The individual strategies above address what graduates can do given current conditions. The more important question is what conditions need to change - and who is responsible for changing them.

Colleges need to treat work experience as a graduation requirement, not an amenity:

Jane Swift and Ryan Craig, writing in US News, argue directly that "colleges that can't reliably deliver internships risk watching students walk away, classrooms empty and revenue evaporate as the market shifts to programs that lead to real jobs." The market pressure is real. If four-year degrees are not reliably producing employable graduates, students and families will route around them. Mandatory co-op programs, embedded internships, and university-created work experiences (like the University of Iowa's journalism student newspapers and Saxbys cafes managed entirely by business students) are not experiments - they are the template.

Employers need to rebuild entry-level pipelines rather than eliminating them:

Yale's Sonnenfeld team poses the sharpest version of the institutional question: "With no entry to the workforce, how will younger people develop the skills and wisdom to lead in the future?" Companies that eliminate all junior roles in exchange for AI tools are trading long-term talent development for short-term cost efficiency. IBM's decision to triple entry-level hiring in targeted fields is not just philanthropic - it is a long-term investment in the senior talent pipeline. Organizations that stop hiring junior people today will have gaps in their mid-level leadership in five years.

Policymakers need to fund the transition:

The mismatch between what colleges produce and what employers require is a policy failure as much as an institutional one. The co-op gap - fewer than 0.5% of American students in formal programs versus near-universal participation in Canada - is not an accident. It reflects funding structures, accreditation incentives, and employer relationships that have not been redesigned for the AI era.

The honest framing:

As Washington Monthly concludes: "That does not mean college has stopped paying off. The data say otherwise. But AI is accelerating a shift that leaves the four-year degree valuable and insufficient at the same time." The problem is not that college is worthless. The problem is that college + diploma is no longer enough. College + diploma + demonstrated professional competency is what the market now requires at entry - and building that combination before graduation is a responsibility that currently falls primarily on students, when it should be shared by institutions and employers.

AI Productivity Statistics 2026
How AI is changing workplace productivity - the data on time saved, ROI, and where the gains are actually landing.

AI Adoption Statistics 2026
Enterprise AI adoption rates across industries - the context for which sectors are automating fastest.

AI Coding Tools Statistics 2026
How Claude Code, GitHub Copilot, and Cursor are changing software development workflows - relevant for CS students specifically.

Claude Code Statistics 2026
The tool that has most directly changed entry-level software development demand - what it can do and what it means for developers.

AI Market Share 2026
The competitive landscape of AI platforms that are reshaping white-collar work.

AI Industry Statistics 2026
Broader industry data on AI investment and the economic forces reshaping the labor market.

Is AI actually replacing entry-level jobs for college graduates?
The honest answer is: it is reducing demand for specific types of entry-level work while not eliminating the job categories entirely. Entry-level job postings are down 35% since early 2023 and unemployment among recent graduates (ages 22-27) is 5.7% - well above the 4.3% rate for all workers. Stanford researchers found a 16% relative employment decline among workers aged 22-25 in AI-exposed roles over three years. At the same time, 3 million white-collar jobs have been added since ChatGPT launched. AI is removing the bottom layer of many job categories while the categories themselves remain or grow.

What majors are most affected by AI and entry-level job displacement?
Counter to conventional wisdom, computer science and computer engineering graduates are among the most affected - with unemployment rates of 7.0% and 7.8% respectively, among the highest of all majors per the Federal Reserve Bank of New York. Software development, customer service, and accounting roles for workers ages 22-25 have seen the steepest relative employment declines. The fields with more resilience are healthcare, cybersecurity, AI engineering, government, and skilled trades - sectors where physical presence, regulatory accountability, or genuine labor shortages create roles that AI cannot substitute for at current capability levels.

What percentage of college graduates are worried about AI taking their jobs?
89% of recent and upcoming college graduates worry that AI could replace entry-level jobs - up from 64% in 2025 - per Monster's 2026 Graduate AI Readiness Report surveying more than 1,000 graduates. Only 36% say college is preparing them to use AI professionally, despite 50% already using AI tools in academic work. 58% feel anxious about using AI tools in future roles, and only 23% of graduates whose fields have been AI-affected received extensive AI training during their studies.

What is "experience creep" and how does it affect college graduates?
Experience creep - a term coined by Laura Ullrich, lead economist at Indeed - describes the pattern where employers demand more prior experience for jobs that once were available to candidates still acquiring it. AI accelerates experience creep because it automates the basic tasks that entry-level jobs traditionally used to develop. When AI tools handle the analytical, writing, and research work that junior employees used to do, employers expect new hires to operate at a higher level immediately. The result: graduates need to demonstrate professional competency before they can get the job that was supposed to develop professional competency.

What are college graduates doing to land jobs despite AI?
The graduates landing jobs in 2026 share consistent traits: they have work experience before graduating (hired at 81.6% vs 40.7% for those without, per ZipRecruiter), they demonstrate genuine AI tool fluency in interviews rather than just awareness, they treat networking as a primary job search method rather than a supplement to applications, and they keep an open mind about what the first role looks like. The diploma is the prerequisite in 2026, not the differentiator. The differentiator is demonstrated professional competency - which increasingly has to be built during college rather than after graduation.

Which sectors are still hiring college graduates at entry-level in 2026?
Healthcare, cybersecurity, AI engineering and deployment, government and public sector, and skilled trades management are actively hiring entry-level graduates in 2026. IBM tripled entry-level hiring specifically in software development, cybersecurity, and AI engineering. Government agencies at federal, state, and local levels continue hiring in social services, policy, infrastructure, and program management. Skilled trades face genuine labor shortages with management-track roles available for graduates willing to start on the operational floor.

What is the underemployment rate for college graduates in 2026?
Approximately 43% of recent college graduates are now classified as underemployed - working in jobs that do not require a college degree or working part-time when full-time is desired - the highest rate since the pandemic peak, per Metaintro's April 2026 labor market analysis. The share of long-term unemployed workers with bachelor's degrees has grown from approximately 1 in 5 a decade ago to approximately 1 in 3. Only 19% of college graduates say it is a "good time to find a quality job" - compared to 35% of those without degrees, a historical reversal.

Is AI taking entry-level jobs from college graduates?
Yes, in a specific and measurable way. Entry-level job postings are down 35% since early 2023. Unemployment among recent college graduates ages 22-27 is 5.7% versus 4.3% for all workers. Stanford researchers found a 16% relative employment decline for workers ages 22-25 in AI-exposed roles. AI is removing the lowest layer of many job categories - the basic analytical, writing, and research tasks that junior employees traditionally handled - without eliminating the categories themselves. 3 million white-collar jobs have been added since ChatGPT launched in 2022, but the growth is concentrated in experienced roles, not entry-level ones.

What percentage of college graduates worry about AI replacing their jobs?
89% of recent and upcoming college graduates worry that AI could replace entry-level jobs, up from 64% in 2025, per Monster's 2026 Graduate AI Readiness Report of 1,000+ graduates. 43% of new graduates are currently underemployed - the highest rate since the pandemic. Only 36% say college is preparing them to use AI professionally despite 50% already using AI tools academically. 69% believe AI skills will give them a hiring advantage.

What jobs are still available for college graduates despite AI?
Healthcare, cybersecurity, AI engineering, government and public sector, and skilled trades management are actively hiring college graduates in 2026. IBM tripled entry-level hiring in software development, cybersecurity, and AI engineering. Government agencies continue hiring in social services, policy, and program management. New AI-specific roles in model evaluation, prompt engineering, AI safety, and LLMOps have emerged as genuine career on-ramps. Sectors requiring physical presence, regulatory accountability, or domain judgment are most insulated from entry-level displacement.

What should college students do to get jobs in an AI economy?
Based on ZipRecruiter's 2026 data and labor economist analysis: get real work experience before graduating - graduates with it are hired at 81.6% versus 40.7% for those without. Build genuine AI tool fluency, not just awareness - employers increasingly test AI use in real time during interviews. Prioritize networking over application portals in a market with 35% fewer entry-level postings. Stay open to adjacent first roles rather than only applying to ideal positions. And translate college activities into professional competency language - clubs, projects, and problem-solving experiences are legitimate professional credentials when framed correctly.

The honest picture of AI and entry-level jobs in 2026 is neither the apocalypse that AI pessimists predicted nor the business-as-usual that AI optimists promised.

Entry-level job postings are down 35%. Graduate unemployment is elevated. Underemployment has reached its highest rate since the pandemic. The long-term unemployed are increasingly holding college degrees. These are real problems for real people, and commencement speakers who try to explain them away with historical examples of technological progress deserve the boos they are getting.

At the same time: white-collar employment overall has grown. The job categories most cited as AI casualties have expanded. IBM is tripling entry-level hiring. Starting salaries for technical majors are rising. The total elimination of professional work has not arrived and does not appear imminent.

The truth is more specific and more actionable than either extreme. AI is removing the bottom layer of many professional roles - the tasks that served as the learning launchpad for new graduates - while leaving or growing the roles above that layer. The on-ramp is narrowing, not disappearing. And the graduates navigating it successfully are the ones who built demonstrated professional competency before they needed it to be demonstrated.

That is partly on graduates. It is more on colleges that charge four-year tuition without reliably producing work-ready graduates. And it is on employers who eliminated junior roles for short-term cost savings while funding the AI tools that will eventually produce the experienced talent gaps they will need to fill.

The Class of 2026 is not facing a broken economy. They are facing a transitional one - where the rules of the first job have changed faster than the institutions that prepare people for first jobs have adapted. Navigating it requires more intentionality, more flexibility, and more proactivity than any prior generation of graduates needed.

The graduates who bring that intentionality will be fine. The institutions that do not adapt will not.

📨 Don't miss tomorrow's edition. Subscribe free to AI Business Weekly and get our 2026 AI Tools Cheat Sheet instantly - bite-sized AI news every morning, zero hype.

Subscribe Free