AI is taking over traditional entry-level tasks, thereby creating a dangerous gap in the talent pipeline. When junior roles disappear, the most important learning phase for young talent is lost—with consequences for companies. Johann Wachs, Managing Director at eightplaces, analyzes why career entry in the AI era requires a new learning and career architecture—and how HR can bridge the gap between algorithms and expertise.
The first job needs to be reinvented – thanks to AI
Job postings for entry-level positions are hard to come by these days. One reason: AI is taking over more and more tasks that new graduates traditionally start with. Companies now need to rethink their entry-level structures – or they’ll lose an entire generation. Young business graduates with master’s degrees looking for work today know this all too well: “We’re looking for someone with more experience” is the feedback they receive—if the rejection isn’t automated right from the start.
Just three years ago, these roles—such as junior consultant, marketing trainee, or junior analyst—were classic entry-level positions. Not anymore. Many of these tasks are now handled by algorithms.
A study by Stepstone confirms what many young people are currently experiencing. In the first quarter of 2025, the number of entry-level jobs posted was actually 45 percent below the five-year average. Administrative and data-processing roles are particularly affected: entry-level positions fell by 56 percent in sales, 50 percent in human resources, and 34 percent in administration.
The reasons for this decline are complex. The economic situation is dampening hiring enthusiasm, and companies are cutting costs first and foremost in entry-level positions. At the same time, however, artificial intelligence is changing job roles: research, data preparation, drafting, and summarizing—tasks that new graduates traditionally start with to gain experience and confidence—are now being handled by AI tools.
What remains are complex, experience-intensive tasks that require expertise and sound judgment. A senior analyst, for example, can immediately recognize when an AI-generated analysis is completely unrealistic. A newcomer to the profession who has never learned to structure raw data on their own sees only the finished result—and has no frame of reference to question it. The learning loop breaks down—and this is precisely where the problem arises for the next generation. They become prompt operators without understanding what is happening behind the scenes.
Three Areas of Action for HR and Leadership
Consequently, companies that fail to act now risk a twofold divide: between generations and between learners and performers. Yet this transformation can be harnessed constructively. The first job doesn’t have to disappear—it “just” needs to be reimagined. Three approaches have proven effective in practice:
1. From Tools to “Decision Cases”: Onboarding in the AI Era
Traditional onboarding starts with software training: “This is how our CRM works, this is how you use Excel, this is how the workflow runs.” That’s no longer enough. Instead, we need to provide context: Why do we use these tools? What decisions do we make with them? Where does AI help—and where doesn’t it?
An example: A publishing house wants to integrate interns. In the first week, a new hire works with experienced colleagues to analyze how AI tools can be used for research and writing—without violating the press code. The learning outcome: The junior staff member learns to approach AI systems critically—for example, by verifying sources and avoiding the use of fake news, rumors, or discriminatory content.
These “decision cases,” in which real-life decision-making scenarios are simulated, bring onboarding into the AI era. The task for new, young employees is therefore not to find the right answer, but to ask the right questions. A senior provides feedback not on the result, but on the thought process.
2. Making experiential knowledge trainable: Tandem models between juniors, seniors, and AI
Any company that views AI as a replacement for junior staff is making a fatal mistake. Instead, AI should function as a third player on the team: junior staff operate the AI, senior staff review the results—and both learn from each other. In a strategy consulting firm, this type of human-AI co-creation could unfold as follows: The junior uses AI to identify patterns in market research data and formulate initial hypotheses. The senior checks whether the patterns are valid or whether the AI is confusing correlations with causality. Then both reflect together: What did the AI overlook? Which questions could it have answered better? What did we learn in the process?
This makes experiential knowledge something that can be practiced, not just observed. The junior doesn’t learn through “shadowing,” but through active participation—with real-time feedback. The senior, in turn, learns to articulate more precisely what matters. Practical pair sessions—short reflections after each AI-supported project—are helpful in this regard. An internal prompt playground, where juniors practice tasks and experienced colleagues provide feedback, reinforces this effect.
3. Learning with AI, not from it: Virtual GPT Coaches
In so-called AI mentoring buddy systems, virtual GPT coaches guide new employees through tasks that require critical thinking: they provide feedback on texts, assist with structuring analyses, or help formulate hypotheses clearly. The goal is not to replace human mentoring, but to expand upon it—with a digital sparring partner who is available at all times and continuously encourages reflection.
AI-augmented learning journeys or micro-learning modules can be used to embed this principle specifically into everyday work, for example through focused, interactive “Critical Thinking with AI” sessions. Questions such as “What would happen if this AI output were incorrect?” or “What assumption lies behind this result?” sharpen judgment.
The first task must be to create an AI-optimized learning system
Because the job profile in an entry-level position is shifting from relatively simple routine tasks toward the targeted use of AI, companies must overhaul the traditional career ladder. New structures must combine the ability to learn with practical experience. HR and leadership should therefore view AI as an opportunity to replace previously ineffective learning structures.
After all, many onboarding processes have been inefficient up to now: junior employees performed monotonous tasks that had little to do with strategic thinking. AI now forces them to question these routines. Today, the goal is to design entry-level positions so that people learn from the very beginning what AI cannot do—namely, create context, tolerate contradictions, and make decisions under uncertainty.
Companies that establish AI-compatible onboarding processes, tandem models, and new mentoring formats now are ensuring their ability to thrive in a world where algorithms are fast—but people set the direction. They gain not only new talent but also innovative strength.
About the Author
Johann Wachs is the Managing Director of eightplaces, an agency specializing in employer branding and HR marketing. Before founding eightplaces, he spent over 25 years in international leadership roles—including at Saatchi, Ogilvy, Grey, and Dentsu—where he was responsible for psychological research on brands and target audiences.
Source: hrjournal.de