New hiring research raises a difficult question for graduates and career changers: what happens when the work that teaches beginners becomes easier to automate?
You keep hearing that employers want experience. But how do you get that experience if fewer employers are willing to hire someone who is still learning?
A new report from one AI company offers a revealing example of that tension, and a reason to rethink how you present yourself when moving into a different field. Below, what the sources report is kept separate from SmartRolePath's own practical recommendations.
Quick answer: AI may change the tasks available to beginners, but evidence from one employer does not show that entry-level jobs are disappearing everywhere. Career changers can start by checking current job requirements, identifying experience that transfers and building an honest work sample that demonstrates their decisions.
Source and scope: Anthropic, "How employment is changing at Anthropic," published October 16, 2026. The report covers Anthropic's own workforce and hiring.
Anthropic reports that junior roles' share of its technical hires fell from 27% to 9%. A separate figure in the same report, the share of junior-level employees in technical departments, fell from 21% to 14% since early 2025. The report also says that, among technical candidates, comments on applicants' autonomy and judgment rose the most in hiring managers' interview assessments.
Our interpretation, not a finding: the case still raises a question that reaches well beyond technology. If software handles more beginner tasks, where will people learn to do the harder work?
Source and scope: Federal Reserve, Beige Book summary, October 14, 2026, which reports anecdotal information from business contacts collected on or before October 5.
The national summary says contacts in five Districts reported that some firms had reduced staffing levels through attrition, delayed hiring, or anticipated less job creation after increasing productivity by investing in AI.
Limitation: this is qualitative evidence from business contacts. It is not a count of jobs eliminated by AI.
Source and scope: U.S. Bureau of Labor Statistics, The Employment Situation, September 2026, released October 2, 2026, covering the U.S. as a whole.
Total nonfarm payroll employment rose by 29,000 in September and the unemployment rate was 4.2%. BLS described both measures as little changed.
Limitation: national employment statistics do not tell us how much AI contributed to hiring decisions, so they cannot establish that AI caused any of the slowdown.
You may need to learn a different tool, industry vocabulary or qualification. That does not erase the experience you already have.
These are possible connections to investigate, not automatic qualifications. Each target role has its own requirements. The useful question is: which parts of this job are new to me, and which parts have I already demonstrated elsewhere?
Our guide to finding the transferable skills you already have can help you name them, and changing careers with no direct experience covers how to approach a field where you have not yet worked. Recent graduates may also want our look at the new-grad job market.
Our recommendation: create a small piece of work that makes your transferable experience visible. A course can help you learn. A work sample lets someone examine what you can do with that learning.
Choose a realistic problem, explain your approach and show the result. Include the limitations and what you would check before using it in a real workplace.
Label practice projects clearly. Never invent clients, employment, testimonials or business results. For licensed, clinical and safety-critical work, a portfolio does not replace required education, supervised practice or credentials.
If you want one place to organize examples like these, a personal career website such as CareerSite can present them together. A website organizes your evidence; it cannot guarantee recruiter attention or search rankings.
| Target direction | Example work sample | What to explain |
|---|---|---|
| Operations coordination | A sample scheduling and handoff process | How you handle delays, competing priorities and escalation |
| Digital design | A website redesign for an invented business | Audience, accessibility, mobile layout and design decisions |
| Customer success | A sample onboarding guide | How you identify confusion and help customers reach a useful outcome |
| Data analysis | An analysis of a public dataset | Cleaning decisions, assumptions and what the numbers cannot establish |
"Comfortable with AI" is difficult for an employer to evaluate on its own. A more useful demonstration explains:
For example, a practice-project description might say:
I used AI to draft three onboarding sequences, checked each against the fictional customer requirements, removed unsupported promises and rewrote the handoff instructions.
That tells an employer more than a list of tools, and it gives you something concrete to discuss in an interview. Use public, invented or properly authorized information. Do not upload confidential employer or customer material to demonstrate your skills.
If you are weighing how AI affects the role you already have, see what to do if AI is changing your job.
This plan is editorial advice from SmartRolePath. It has not been tested as an intervention, and it does not promise interviews or job offers.
Review ten recent postings from identifiable employers. Record the recurring tasks, required qualifications, location restrictions and experience expectations. Treat this as a small research sample, not a complete picture of the market, and confirm that postings remain active on the employers' own websites.
Choose a problem you can address with your existing experience plus one manageable new skill. Keep the scope small enough to finish. A clear, thoughtful example is more useful than an ambitious project you cannot explain.
Ask someone familiar with the work: "What would make this more credible or useful in your workplace?" Listen for missing requirements and incorrect assumptions.
Add a short case study to your résumé, LinkedIn profile or personal career website. Explain the problem, your contribution, your decisions and the result.
Connect your previous experience and work sample to that employer's stated requirements. Do not claim that a practice project equals years of professional experience. Use it to show preparation, curiosity and your ability to discuss the work intelligently.
Applicants cannot repair a weakened training pipeline by themselves. Employers should explain how newcomers will learn, receive feedback and progress into greater responsibility. Asking a company how it develops new hires is a reasonable interview question.
If an organization automates beginner tasks, it still needs a plan for developing the people who will eventually handle its complex decisions.
You cannot control whether an employer opens a junior position. You can make it easier to understand what you already know, what you are learning and how you approach unfamiliar problems. Start with one realistic direction, check its requirements and build one honest example.
SmartRolePath's career analysis uses AI to compare your background, interests and goals with possible career paths. Treat the results as a starting point: check each suggested path against actual employer requirements before committing to training or a career move.
Explore career paths that build on your experience
Sources, each checked against the linked page on October 19, 2026:
Limitations:
Editorial note: this article combines reporting from the linked sources with SmartRolePath's practical recommendations. Statements labeled as recommendations or interpretation are ours.
No source reviewed for this article shows that. Anthropic reports that junior roles' share of its technical hires fell from 27% to 9%, but that describes one company's hiring mix, not a national trend. The report describes its technical hiring overall as continuing to grow in absolute terms, which is not a statement about junior hiring. The Federal Reserve's October 14, 2026 Beige Book summary relays qualitative reports from business contacts in five Districts about slower hiring or staffing after AI investment, not a count of jobs lost. BLS national employment figures do not show how much AI contributed to hiring decisions.
It is a signal worth taking seriously, not a forecast for your field. Anthropic's figures describe its own hiring, so a practical step is to check current requirements for the role you want, identify experience that transfers, and build an honest work sample that shows your decisions. The report also found more interviewer comments on applicants' autonomy and judgment among technical candidates, which is one reason we suggest explaining your choices and not only listing tools.
Create a small, realistic work sample, label it clearly as a practice project, and explain your approach, your decisions and its limits. If you use AI, say what you asked it to do, what you checked, and what you corrected or rejected. Use public, invented or properly authorized information, and never invent clients, employment, testimonials or results.
No. A portfolio can show how you work and that you have prepared, but it does not replace required education, licenses, supervised practice or credentials, and it should not be presented as equal to years of professional experience. In licensed, clinical and safety-critical fields, those requirements still apply.