# How is AI changing entry-level jobs?

Date: 2025-08-23 · Author: Robert Scales · https://guidetoartificialintelligence.com/en/learn/ai-changing-entry-level-jobs/

**Answer:** The evidence so far is mixed. Hiring of new graduates has dropped sharply in tech, and AI is one of several suspected reasons, but broad studies of US labor data this month find no clear sign that AI-exposed workers, including young ones, are losing jobs faster than others.

Many recent graduates are finding the first job harder to land than their older siblings did, and AI is often named as the reason. Chatbots now draft emails, summarize documents and write simple code, which are the kinds of tasks junior staff have traditionally learned on. It is a reasonable worry. The research, though, shows a more complicated picture than the headlines.

Two recent analyses are worth reading side by side: one looks at hiring in the tech industry, the other at the whole US labor market. They answer different questions, and both come with limits.

## What the hiring data shows

In May, the venture firm SignalFire published its 2025 report on tech talent. At large tech companies, new graduates made up about 7% of hires in 2024. Hiring of new graduates there was down 25% from 2023 and more than half below 2019 levels. At startups, new graduates were under 6% of hires.

SignalFire does not blame AI alone. It lists smaller funding rounds, leaner teams and fewer graduate programs alongside AI tools that can now handle some routine junior work. It also notes that experienced engineers were still in demand, and that AI had not wiped out whole job categories. In short: fewer doors for beginners in tech, with several causes mixed together.

The limits matter. This is tech hiring, not the economy as a whole, and a drop in hiring after the boom years of 2021 and 2022 would be expected even without AI.

## What the broader labor data shows

On August 10, the Economic Innovation Group, a policy research organization, published an analysis that looks for AI's effect across US workers. To decide who is "exposed," it used five different measures from four research papers. An **exposure** score estimates how many of a job's tasks AI could perform; it says nothing about whether AI is actually being used in that job.

The results were mostly negative, in the scientific sense of finding no effect. Between 2022 and early 2025, unemployment rose 0.30 percentage points for the most AI-exposed fifth of workers and 0.94 points for the least exposed. Exposed workers were not leaving the workforce or switching careers more than before. Most relevant here, unemployment has been rising for young workers and recent graduates whether their fields are exposed to AI or not. Two of the five measures did show a small relative rise for exposed workers, about 0.2 to 0.3 points.

The authors are careful about what this means. Exposure measures can't prove cause, and only about 9% of US businesses told the Census Bureau they use AI to produce goods or services, so effects may simply not have arrived yet.

## Reading the two together

| Question | Tech hiring data | Broad labor data |
|---|---|---|
| Are beginners hired less? | Yes, sharply, in tech | Young workers' unemployment is rising |
| Is AI the cause? | One of several reasons | No clear sign so far |
| Main limit | One industry, many causes | Exposure is not the same as use |

The fair conclusion for August 2025: the entry-level market is tougher, especially in tech, and AI may be part of that, but the evidence does not yet show AI pushing young people out of work on a large scale. New studies are appearing quickly, and each one deserves the same questions: who was measured, how was AI exposure defined, and could something else explain the pattern?

## What can new graduates do about it?

Treat the uncertainty as a reason to prepare, not to panic. The practical goal is to show an employer that you can do useful work from the first week, including work done well with AI tools. SignalFire's advice points the same way: employers want proof, so build it through projects, open-source contributions or freelance work, and learn to find and fix the mistakes AI makes.

Here is a worked plan for one season, using a marketing graduate as the example. Swap in your own field.

| Step | What to do | What you can show |
|---|---|---|
| 1. Pick tasks | List five routine tasks from real job ads in your field (e.g., writing product descriptions, reporting campaign results) | A short list tied to actual postings |
| 2. Do them with AI | Complete each task using a chatbot, then fix what it got wrong | Before-and-after drafts with your edits marked |
| 3. Add judgment | For each task, write two lines on what the AI missed and why it mattered | Evidence that you can review, not just generate |
| 4. Make it real | Offer one task to a local nonprofit or small business for free or a small fee | A reference and a real result |
| 5. Tell the story | Put the best two examples in a one-page portfolio | Something concrete to discuss in interviews |

> **Try it:** Find one entry-level job ad in your field today. Copy three tasks from it, do one of them with an AI tool this week, and write down every correction you had to make. That list is the start of your portfolio.

Two more habits help. Keep an eye on the parts of a job that need people: talking with clients, making judgment calls, taking responsibility for a result. And widen the search beyond the largest companies, since the hiring drop has been steepest there. None of this guarantees a job, but it answers the question employers are now asking: what can you do that the tools can't do alone?

## Sources

- AI and Jobs: The Final Word (Until the Next One): https://eig.org/ai-and-jobs-the-final-word/
- The SignalFire State of Tech Talent Report - 2025: https://www.signalfire.com/blog/signalfire-state-of-talent-report-2025