The AI Excuse Is Covering Up the Destruction of Entry-Level Careers (And Companies Will Regret It)

Last week, I wrote about how Trump’s policy chaos, DEI rollbacks, and economic uncertainty are destroying entry-level careers. Now I’m mad about something else (in addition to, not in replacement of, I’m still mad about that other stuff.

Companies aren’t just cutting entry-level positions because of tariff uncertainty or immigration restrictions. They’re using AI as the perfect cover story to get rid of roles they never intended to keep, while pretending it’s about innovation rather than incompetent jackassery.

Quick note to point out that I use AI regularly and while I am conflicted about proper ethics and energy use and all the other conflict stuff, I am what you might call a cautious advocate of responsible AI usage.

And the data? It’s damning.

Companies spent billions on AI and got nothing back—now they’re cutting entry-level jobs and blaming robots that don’t even work.

The AI Lie: 95% ROI of Exactly Nothing

Here’s what MIT researchers just discovered: 95% of companies investing in AI are getting exactly zero return on investment. These companies spent between $30-40 billion on AI initiatives, and 95% of them saw no measurable impact on profits.

Zero. Zilch. Nothing.

Yet somehow, 180,000 tech workers lost their jobs in 2025, with companies pointing to AI as the reason.

The math doesn’t math.

According to the New York Federal Reserve, only 1% of service firms reported laying off workers due to AI in the past six months — down from 10% last year. But layoffs are at record highs, averaging 489 jobs lost per day.

So what’s the story, morning glory? Companies are using “AI transformation” as PR cover for the cost-cutting they were gonna do anyway. And the first positions to disappear? Yeah, you guessed it. Sorry, little Jimmy, your entry-level role went into some shareholder’s pocket. But sure, blame AI.

The Entry-Level Massacre (Now With Data)

Remember when I wrote about how 85% of job seekers say the market is bad and people are submitting 91 applications for 6 responses? Here’s why that’s happening.

Dario Amodei, CEO of Anthropic (an AI company, ironically), told Axios that AI could eliminate half of all entry-level white-collar jobs. Between October 2022 and July 2025, workers aged 22-25 in “highly exposed” fields saw a 13% relative decline in job opportunities.

These aren’t senior positions being cut. It’s junior analysts, entry-level developers, assistant roles — the traditional starting points for careers. The positions where you learn organizational culture, develop skills, and build the relationships that make you successful in more senior roles.

But here’s the twist that connects everything I wrote last week: Companies aren’t actually replacing these roles with functional AI. They’re just not filling the positions when people leave.

“Soft attrition,” as the MIT report calls it. No dramatic layoffs. Just a quiet erosion of opportunities.

And when you layer in the tariff chaos, making workforce planning impossible, the H-1B restrictions that eliminate pathways for early-career tech workers, and the DEI rollbacks that remove sponsorship programs that helped people advance… you get a perfect storm where entry-level careers are being systematically eliminated while companies pretend it’s about innovation.

The Companies Getting Caught

Let’s name names, because the specifics matter.

Salesforce slashed 4,000 customer support jobs in September, with CEO Marc Benioff claiming AI can do “50% of the work.” But Salesforce’s own spokesperson later admitted they “redeployed hundreds of employees into other areas.” Translation: We moved people around, not AI.

Klarna became the poster child for AI replacement after cutting 40% of its workforce. CEO Sebastian Siemiatkowski initially bragged about AI doing the work of 700 customer service agents. But in a recent damage-control post, he clarified: “We have made 0 layoffs due to AI.” The cuts were due to “natural attrition” after they stopped hiring in 2023.

The company everyone points to as proof AI kills jobs just admitted AI didn’t kill those jobs.

Accenture announced it was “exiting” 11,000 employees who couldn’t be “reskilled” on AI. Except the company reported 7% revenue growth while making these cuts. They’re profitable. They just decided they needed fewer humans.

And here’s where it connects to everything: IBM and Klarna both reversed course on AI customer service after discovering the technology couldn’t actually handle the complexity of human interactions. They spent millions. Cut thousands of jobs. Then had to start rehiring because the AI failed.

Sound familiar? It should. It’s the same pattern I described before: companies making shortsighted decisions during chaos, using whatever excuse sounds good in the press release, without understanding they’re destroying the talent pipelines they’ll need in five years.

Why “AI Transformation” Is the Perfect Scapegoat

Fabian Stephany, assistant professor of AI and Work at the Oxford Internet Institute, doesn’t mince words: Companies are “scapegoating” AI to cover up old-fashioned cost-cutting.

Think about it from a CEO’s perspective:

Option A: Tell investors you’re laying off 5,000 people because you overhired during the pandemic, miscalculated your growth projections, and need to boost quarterly earnings.

Option B: Tell investors you’re “strategically realigning toward an AI-first future” and cutting 5,000 jobs as part of your “innovative digital transformation.”

Which sounds better in a press release? Which one makes your stock price go up?

Professor Stephany nails it: “Companies can now position themselves at the frontier of AI technology to appear innovative and competitive, and simultaneously conceal the real reasons for layoffs.”

And when you’re operating in an environment where tariffs change weekly, immigration policy targets entry-level pathways, and economic uncertainty makes long-term planning impossible… “AI transformation” becomes the perfect excuse to eliminate positions you’re too damn scared to fill anyway.

The Compounding Disaster

Here’s where everything converges with the AI excuse:

330,000 women left the workforce by August 2025. Companies are cutting entry-level positions and blaming AI. One in five companies cut diversity staff and resources over the past 12 months. Tariff chaos makes planning impossible. H-1B restrictions target exactly the entry-level tech roles that build pipelines.

And through it all, companies point at AI and say “the robots made us do it.”

Except the robots aren’t working y’all. At least not the way these people expected it to. That MIT study found that only 5% of AI pilots extract measurable value (IDK what that means, but), most tools “fail to contribute to profits due to brittle workflows and misalignment with operations,” and internal AI builds succeed only 33% of the time.

So companies are:

• Eliminating entry-level positions (destroying talent pipelines)
• Blaming AI (which isn’t actually working)
• Cutting DEI programs (removing pathways for women and people of color)
• Operating in policy chaos (making planning impossible)
• Pretending it’s innovation (when it’s really incompetence)

And in five years, heck, in TWO years, when they’re desperate for mid-level talent and wondering where their pipeline went, they’ll act surprised.

The AI revolution isn’t taking your job—the AI excuse might. And companies are destroying talent pipelines while pretending to innovate.

What Yale and MIT Actually Found

The research is clear, and it contradicts everything companies are claiming.

Yale University’s Budget Lab looked at U.S. labor data from November 2022 – July 2025 using a “dissimilarity index” to measure how much the job market has shifted since ChatGPT rode into town. They compared it to other major tech disruptions like computers and the internet…you know, those quaint old revolutions.

tl;dr The AI disruption everyone’s freaking out about isn’t happening on a mass scale.

MIT’s “The GenAI Divide: State of AI in Business 2025” analyzed 300 public AI deployments and interviewed 52 executives. They found:

• Only 5% of AI pilots extract measurable value
• Companies waste money on AI for sales and marketing (low ROI) instead of back-office automation (high ROI)
• Internal AI builds succeed only 33% of the time vs. 67% for purchased solutions

Translation: Companies have no idea what they’re doing with AI, but they’re happy to use it as cover for layoffs they planned anyway.

The Pandemic Overhiring Nobody Wants to Admit

Here’s what companies really don’t want to talk about: Many of them significantly overhired during COVID-19.

When everyone went digital in 2020-2021, tech companies went on hiring sprees. Demand seemed infinite. Growth projections looked insane. VCs threw money at anything with a .com.

Now? Reality check. Demand normalized. Investor patience evaporated. Interest rates went up. And suddenly all those extra employees hired during the boom years look like “inefficiencies.”

Companies like Duolingo, Klarna, and dozens of others are doing what Stephany calls “market clearance” — firing people they never should have hired in the first place.

But saying “we screwed up our hiring forecasts in 2021” makes you look incompetent. Saying “we’re embracing AI” makes you look like a visionary.

Same layoffs. Different spin.

And when you add Trump’s tariff chaos making it impossible to forecast 60 days out, immigration restrictions targeting entry-level pathways, and economic uncertainty making every decision feel high-stakes… the AI excuse becomes irresistible.

The Data That Connects Everything

Let me lay out the full picture:

• 85% of job seekers say the market is bad
• Average job search: 9 months
• 91 applications for just 6 responses
• 330,000 women left the workforce by August
• ADP showed private sector employment fell 32,000 in November
• Women are 13% less likely to be promoted at entry level

From the AI research:

• 95% of companies get zero ROI from AI investments
• 180,000 tech workers lost jobs in 2025
• Only 1% of service firms actually laid off workers due to AI
• Workers aged 22-25 saw 13% decline in job opportunities
• Only 5% of AI pilots extract measurable value

These aren’t separate crises. They’re one compound disaster where:

• Policy chaos makes planning impossible
• Companies use AI as cover for cuts they wanted anyway
• Entry-level positions disappear (destroying pipelines)
• Women and people of color lose ground
• DEI programs get eliminated
• The job market becomes dysfunctional for everyone

And GDP grows 3.8% while economists wonder why workers aren’t celebrating.

What This Means for HR Tech Companies

Let me be very direct about what this means for my industry.

Your clients are about to make catastrophically bad decisions. They’re going to:

• Eliminate entry-level positions and call it “AI transformation”
• Cut development programs and call it “efficiency”
• Roll back DEI initiatives and call it “returning to merit”
• Use AI tools that don’t work and pretend they do
• Destroy talent pipelines and act surprised in five years

And your job — our job — is to either enable this stupidity or push back against it.

The HR tech solutions that will matter in 2026 are:

• Internal mobility platforms that help companies maximize existing talent when they can’t hire
• Retention technology that becomes critical when the pipeline dries up
• Skills development tools that help companies actually reskill people (not just say they will)
• DEI analytics that prove the business case when leadership wants to cut programs
• Workforce planning tools that account for policy chaos and uncertainty

The competitive advantage is moving from “we can help you hire faster” to “we can help you survive when hiring becomes impossible.”

In 2030, companies will be desperate for mid-level talent. They’ll be scrambling to rebuild the entry-level programs they eliminated. They’ll be trying to restart the DEI initiatives they cut. They’ll realize that the AI tools they’ve spent billions on don’t actually replace human judgment, relationship-building, or institutional knowledge.

And they’ll discover that the workers they pushed out remember exactly which companies abandoned them during chaos and which companies fought to keep them.

The AI bubble will adjust (it always does). The tariff chaos will eventually stabilize (one way or another). The policy uncertainty will settle (at some point).

But the trust you didn’t build doesn’t come back so quickly.

What Workers Can Actually Do

Alright, enough analysis. Here’s what matters:

1) Your job probably isn’t being replaced by AI. It might be eliminated by cost-cutting that blames AI. That’s a crucial distinction.

B) Document everything. Track metrics. Build relationships across departments. Make yourself visibly valuable. Being good at your job isn’t enough — you need to be costly to replace.

Third: Learn AI tools, but not because they’ll replace you. Learn them because they’re productivity multipliers. A worker who knows how to use AI tools to do their job faster is worth keeping.

4: Watch the money, not the marketing. If your company announces “AI transformation” but isn’t investing in training, infrastructure, or implementation — it’s PR. Real AI initiatives involve months of pilots and employee training. If they skip straight to layoffs, it was never about AI.

D: Know your worth outside your current company. Loyalty means nothing anymore. Your best protection is knowing you can land somewhere else.

Sixth (and this is new): If you’re early in your career, focus on building skills that AI genuinely can’t replicate: complex relationship management, cross-functional coordination, judgment calls that require organizational context, and work that involves navigating ambiguity. The AI can’t do those things (yet, and maybe ever).

The Bottom Line

Companies spent billions on AI and got almost nothing back. Now they’re cutting costs and blaming the robots. Entry-level positions are disappearing. Women and people of color are losing ground. DEI programs are being eliminated. Policy chaos makes planning impossible.

And executives are collecting bonuses for “strategic transformation,” while 36% of successful job candidates admit to lying just to get through the door because the market has become that dysfunctional.

The AI revolution isn’t taking your job. But the AI excuse might.

And in five years, when companies realize they’ve destroyed their talent pipelines while pretending to be innovative, it’ll be too late to fix it.

The question is: Are HR tech companies going to enable this disaster, or are we going to be the ones saying “this is a terrible idea and here’s the data proving it”?

Because right now, we’re the only ones with a clear view of the compounding crisis. And if we don’t speak up, nobody will.